Regulation & Compliance Archives - My Value Solutions

Will AI Replace Accountants and Financial Analysts? What’s Actually Changing in 2026

MyValueSolution — Wed, 22 Apr 2026 11:22:26 +0000

85%

CPA Exam (GPT-4, Auditing)

75%

Current CPAs Eligible to Retire

300K+

Left the Field, 2019–2024

44%

Of Adopters Use GenAI Daily

The anxiety is loud. The reality is quieter.

Scroll through any accountant’s or financial analyst’s professional feed for ten minutes and you will be told, with considerable confidence, that the profession is about to end. The evidence cited is usually impressive: large language models passing professional exams, software handling thousands of invoices with no human involvement, audit engagements compressed from weeks to days. The conclusion offered is usually the same. Humans are being replaced.

The conclusion is wrong, or at least substantially oversimplified. What is actually happening in 2026 is a structural transformation of the profession rather than a replacement of it. Specific tasks are being automated — aggressively, at scale, in ways that would have seemed implausible five years ago. Specific jobs are changing beyond recognition. But the demand for skilled accountants, financial analysts, valuation professionals, and audit partners is not falling. In many markets, it is rising.

Understanding the difference between “task replacement” and “job replacement” is the single most useful mental frame for reading the current moment. This piece walks through what is actually happening, what the data says about adoption and outcomes, where AI genuinely wins, where it doesn’t, and what finance professionals should actually do in response.

Where the replacement panic came from.

The fear has a specific origin, and it is worth naming. In early 2023, an earlier generation of large language model — GPT-3.5 — was tested on the US Certified Public Accountant exam. It failed, with an average score of 48%. Eighteen months later, GPT-4 averaged 85.1% across all four sections, including a 91.5% on Auditing and Attestation. The jump in capability was not linear; it was step-change. For a profession built on technical knowledge and pattern recognition, the implication felt obvious.

Compounding the exam result was a broader perception problem. To someone outside the profession, accounting looks like data entry, arithmetic, and spreadsheet manipulation. AI is good at all three. The inference is simple: if machines can do the work, the people currently doing it are in trouble.

What the inference misses is what professional accounting actually involves. Exam-passing tests technical knowledge retrieval and structured problem-solving — both areas where large language models are now genuinely strong. But professional practice involves judgement, context, interpretation, accountability, and relationship management. Those are areas where the same models remain substantially weaker — and where the weakness matters the most, because the cost of an error is borne by a human professional who signed off on it.

Table 1 — LLM Performance on the CPA Exam, by Section
Section	GPT-3.5 (2023)	GPT-4 (2024)	Change
Auditing & Attestation (AUD)	52%	91.5%	+39.5 pts
Business Environment (BEC)	58%	88.0%	+30.0 pts
Financial Accounting (FAR)	41%	82.3%	+41.3 pts
Regulation (REG)	41%	78.5%	+37.5 pts
Overall Average	48%	85.1%	+37.1 pts

Eighteen months of capability change. Whether that change will continue at the same pace through the next cycle is the question the profession cannot yet answer.

What AI is actually automating — the “boring stuff.”

A researcher at Stanford’s Graduate School of Business recently described AI’s role in accounting as removing the “drudgery” from the profession. It is a useful framing. The work that is being automated — aggressively, successfully, at scale — is the repetitive, rules-based, high-volume work that accounted for a substantial portion of a junior accountant’s week until very recently.

The four most mature categories of automation, in 2026, look like this.

01 · MATURE

Transaction capture and classification

OCR engines now read receipts and invoices with close to 100% accuracy on standard document layouts, extracting vendor, amount, date, and line-item data automatically. Machine-learning classifiers then categorise the transactions against the chart of accounts. Corrections from the human reviewer feed back into the model, so accuracy improves over time. The practical result: thousands of transactions processed per month with no manual data entry.

02 · MATURE

Bank reconciliation and three-way matching

Reconciliation has become the single most automated workflow in modern accounting platforms. AI matches invoices to purchase orders and delivery confirmations, runs bank reconciliations in minutes rather than days, and flags exceptions for human review. For firms doing monthly bookkeeping for recurring clients, this alone has cut engagement labour by 40 to 60 percent.

03 · PRODUCTION-GRADE

First-pass analysis and variance commentary

Generative AI is now routinely used to produce first drafts of variance analyses, management commentaries, and trend explanations from P&L and balance-sheet data. The outputs are not final; they are a starting point that a human analyst edits and signs off on. Time savings on a typical monthly commentary run 40 to 60 percent.

04 · PRODUCTION-GRADE

100% population review in audit

The Big Four have quietly rebuilt their audit platforms around the ability to scan entire transaction populations rather than samples. An auditor who previously tested 50 of 50,000 invoices now reviews all 50,000 and spends their time on the ones flagged as anomalous. The qualitative shift: less time on testing, more time on interpretation.

“

AI is not replacing accountants. It is replacing the specific tasks accountants were already hoping someone else would do.

Editorial Assessment

The work that is not going anywhere.

Every honest assessment of AI in professional finance has to address the opposite side of the ledger: the work the technology cannot do, will not do soon, and in several cases structurally should not do. Four categories stand out.

Professional judgement under conditions of ambiguity. Revenue recognition under ASC 606 depends on understanding contract intent. Fair-value measurement relies on judgement about assumptions and market conditions. Lease classification, materiality thresholds, impairment indicators, and contingent liabilities all require contextual assessment that a trained professional applies and is accountable for. AI can surface the relevant facts. It cannot form the opinion.

Accountability that holds up in front of regulators. The US Securities and Exchange Commission holds management responsible for the accuracy of financial statements. Auditing standards require professional scepticism and documented judgement. Current regulatory frameworks do not permit an AI system to sign off on a financial statement or an audit opinion. Someone qualified must. That requirement is not changing in 2026, and probably not in 2030.

Hallucination risk on high-stakes outputs. Large language models produce confident-sounding outputs that are sometimes wrong — citing cases that do not exist, inventing financial ratios, misstating tax-code sections. The probability of error is low on well-defined tasks with abundant training data. The cost of an undetected error on a tax memo, a valuation assumption, or a financial-statement disclosure is disproportionately high. Every serious deployment pairs AI output with mandatory human review for exactly this reason.

The human side of advisory and valuation work. Client relationships are built on trust, interpretation, and the ability to read between the lines of what a client is and is not saying. Tax planning depends on personal goals and risk tolerance. Business valuation engagements hinge on management assumptions, industry context, and judgement about comparability. These are the areas where the best accountants and analysts are actually paid — and where AI remains an assistant, not a replacement.

Table 2 — The Task-Replacement Matrix, 2026
Activity	Automation Status	Human Role
Receipt and invoice capture	Mature	Exception review only
Transaction classification	Mature	Correct and retrain
Bank reconciliation	Mature	Handle anomalies
Three-way matching (AP)	Mature	Approve exceptions
Variance commentary drafting	Production-grade	Edit and contextualise
Tax research drafting	Production-grade	Verify citations, apply context
Audit anomaly detection	Production-grade	Interpret the flags
Financial statement preparation	Partial	Prepare, review, sign off
Valuation assumption setting	Limited	Set, justify, defend
Accounting-standards interpretation	Low	Primary responsibility
Client advisory conversations	Minimal	Primary responsibility
Audit opinion / sign-off	Not permitted	Professional and regulatory requirement

The pattern is consistent across categories: the lower the cost of error and the higher the volume, the more automation is present. The inverse is also true.

The profession’s real problem isn’t AI. It’s demographics.

The replacement-panic narrative misses the actual crisis facing the profession. Between 2019 and 2024, more than 300,000 accountants left the field. Approximately 75% of current practising CPAs in the United States are eligible to retire within the next decade. Accounting-programme enrolment at universities has been declining for years. New entrants to the profession fall short of demand by somewhere in the range of 75,000 heads per year.

The reasons are not mysterious. Audit and tax busy seasons routinely push weekly hours to 60 or 80 over sustained multi-month periods. Burnout is endemic. Pay for junior staff has not kept pace with the opportunity cost of alternative career paths in technology or finance. The combination of long hours, slow pay progression, and rigid credentialing requirements has made the profession substantially less attractive to people entering the workforce.

Against this backdrop, AI is not a threat. It is a lifeline. Every hour of routine transaction processing that a machine handles is an hour a human accountant does not spend on it — an hour that can be spent on higher-value work, on client relationships, or on simply going home. Firms that have invested early in automation are reporting measurably lower attrition, shorter busy seasons, and better retention of mid-career staff. The technology is solving the people problem, not creating it.

Table 3 — The Accountant Supply Gap, Key Metrics
Metric	Figure	Direction
Accountants who left the field, 2019–2024	300,000+	↓ supply
Current CPAs eligible to retire within a decade	~75%	↓ supply
Estimated annual shortfall of new US entrants	~75,000	↓ supply
Peak busy-season weekly hours	60–80	↑ burnout
Engagements firms report turning away due to staffing	Rising	↑ unmet demand
Bookkeeping labour savings from automation	40–60%	↑ capacity

From scorekeeper to strategic partner.

For most of the profession’s modern history, accounting has been fundamentally backward-looking. The job was to record what happened, close the books, produce the report, ensure compliance, and file it on time. The historical stereotype — careful, precise, somewhat unglamorous — reflected the work itself.

The role emerging in 2026 looks different. Automation handles the backward-looking reporting efficiently enough that human professionals are pulled further up the value chain. Financial planning and analysis matters more than bank reconciliation. Cash-flow forecasting matters more than data entry. Tax optimisation matters more than tax compliance. Valuation work is becoming more common as a mid-market service because the tooling makes it economically viable at engagement sizes that previously did not support it.

This transition is happening worldwide, with local variations. In the Nordic markets, where digital accounting infrastructure has been mature for more than a decade, finance professionals have been operating as strategic advisors for some time. Swedish firms — for instance a redovisningsbyrå i Stockholm serving tech-startup and consultancy clients — have typically offered integrated bookkeeping, tax planning, and advisory work under one roof, because the digital foundations were already there. Markets where firms are still migrating from desktop software to cloud platforms are reaching that same operating model two to five years later. The direction of travel is uniform; the timing varies by geography.

The three skills that now matter most.

The changing role brings a changing skill set. Three capabilities are emerging as the ones that separate accountants and analysts who are thriving in the new environment from the ones who are struggling.

Working effectively with AI

Prompting well, using the right tool for the right task, knowing when to override an AI suggestion and when to accept it. This is becoming the 2026 equivalent of Excel fluency in the 1990s — an expected baseline skill rather than a differentiator.

Critical validation of AI outputs

The ability to look at an AI-drafted memo or variance analysis and spot the error — the wrong citation, the hallucinated ratio, the misclassified transaction. This is where professional experience shows, and it is also where junior staff who skip this step create the most liability for their firms.

Communication, translation, and client judgement

Explaining an insight, translating numbers into business decisions, advising a client through a nuanced tax or valuation question. The soft skills that were always undervalued in a technical profession are now the ones that define the professionals who command premium rates.

The accountant who loses their job to AI in 2026 loses it to a younger accountant who knows how to use AI better — not to the machine itself.

The Real Competitive Risk

The hybrid future is already here.

The future of accounting and financial analysis is not humans or AI. It is both, operating together, with a clearer division of labour than the profession has had for fifty years. AI handles the high-volume, rules-based, repetitive work. Humans handle judgement, interpretation, accountability, relationships, and the sign-off on anything that matters.

The firms that understand this division are already pulling ahead of the ones that do not. Their engagements are more profitable, their staff are less burned out, and their client relationships are measurably stronger. The firms that are still treating AI as either a threat to be resisted or a panacea to be over-deployed are struggling in both directions — unable to capture the efficiency gains, and also unable to maintain the quality standards the profession requires.

The “robo-accountant” narrative was always a misread of what was happening. The actual transformation is both less dramatic and more consequential: a profession that has been remarkably stable in its operating model for generations is being quietly rewired around a new set of tools. The professionals who adapt will thrive. The ones who do not will be outcompeted by the ones who do — not by the AI, but by their colleagues who learned to use it first.

Frequently Asked Questions

Fifteen questions on the future of the profession.

Will AI replace accountants in the next decade?

No. It will restructure what accountants do rather than eliminate the profession. Routine, rules-based, high-volume work will increasingly be automated. Judgement-heavy work — advisory, complex tax, audit sign-off, valuation — will remain firmly with qualified humans.

If AI can pass the CPA exam, why can’t it replace CPAs?

Exam-passing tests technical knowledge retrieval and structured problem-solving. Professional practice requires judgement under ambiguity, accountability to regulators, context-specific interpretation, and client relationship management. The exam is a filter for entry, not a measure of the job.

Which accounting tasks are already fully automated?

Transaction capture via OCR, transaction classification, bank reconciliation, and three-way matching in accounts payable. These workflows now run with minimal human involvement in firms that have deployed modern cloud-accounting platforms.

Which tasks remain firmly with humans?

Professional judgement on accounting standards, signing off on financial statements and audit opinions, setting valuation assumptions, interpreting regulatory frameworks, and advising clients. Regulatory frameworks require a qualified human in the loop for each of these.

How much faster can AI-enabled firms close the books?

Practitioner-level research suggests 7 to 8 days faster on monthly close cycles compared to traditional methods, with approximately 8 to 9% less time on routine back-office processing. The time recovered tends to flow into higher-value work rather than into additional bookkeeping volume.

What is the real risk to individual accountants?

Being outcompeted by colleagues who adopted AI tools earlier. The professionals losing market share in 2026 are not losing it to machines; they are losing it to accountants who know how to use AI effectively and price their engagements accordingly.

Is the CPA shortage making AI adoption faster?

Yes, dramatically. With approximately 300,000 accountants leaving the field between 2019 and 2024 and roughly 75% of current CPAs eligible to retire, firms face a binary choice: automate or shrink. Most are choosing to automate, which has made adoption noticeably faster in the last 18 months.

What is AI “hallucination” and why does it matter in accounting?

Hallucination is when an AI system produces confident-sounding output that is factually wrong — citing tax-code sections that do not exist, inventing financial ratios, misstating figures. In accounting, where outputs need to be 100% accurate and are subject to regulatory review, hallucination risk is the reason every serious deployment pairs AI drafting with mandatory human verification.

Are the Big Four ahead of smaller firms on AI?

On proprietary tooling, yes — they have invested billions in internal platforms that mid-market firms cannot match. On directional capability, the gap is much smaller, because the commercial tools available to mid-market firms now offer most of the same workflows. The remaining advantage is integration across the engagement stack.

What skills should accounting students develop now?

Three, in priority order: effective use of AI tools, critical validation of AI outputs, and communication and client judgement. Traditional technical accounting skills remain essential baselines, but they are no longer sufficient for the roles that will be most valuable in five years.

Will senior accountants or junior accountants be affected more?

Junior roles are being affected first, because that is where the most automatable work sits. But senior staff who cannot use AI effectively will find themselves outcompeted by senior staff who can. The disruption reaches up the pyramid; it does not stay at the bottom.

How does valuation work specifically change with AI?

Data gathering, comparable-company analysis, and initial DCF modelling all become substantially faster. But the critical work — choosing the right comparable set, setting discount-rate assumptions, defending the valuation in front of a counterparty — remains firmly with the analyst. AI is making valuation services viable at smaller engagement sizes, not replacing the analyst on the engagement.

What’s the single biggest risk of ignoring AI adoption?

Competitive repricing. Clients who can get a similar engagement done faster and cheaper by a firm using AI will eventually move. The firms that treat automation as optional are operating on borrowed time in the categories where AI handles the work well.

Will the pricing model of accounting engagements change?

It already is. Fixed-fee and value-based pricing are replacing hourly billing in many categories, because the labour hours inside a given engagement are falling. Firms still running pure hourly-billed compliance work are gradually losing share to firms offering fixed-fee, AI-enabled alternatives.

What should a managing partner do on Monday?

Three things. First, find out what AI tools staff are already using without formal policy — the answer will surprise you. Second, pick one workflow with a known bottleneck and run a disciplined pilot. Third, write a one-page AI policy covering client data, human review, and disclosure. The worst governance posture is the one you do not have at all.

Editor’s Note

This analysis draws on published professional-services research, CPA-exam performance data, and practitioner-level observations from firms across multiple markets. MyValue Solutions is editorially independent and not affiliated with any of the professional-services firms, software vendors, or research organisations referenced in this piece. This content represents editorial assessment for information purposes only and should not be construed as investment, tax, or legal advice.

The post Will AI Replace Accountants and Financial Analysts? What’s Actually Changing in 2026 appeared first on My Value Solutions.

How Do I Actually Lower My AI Spend? A Practical Guide for Finance Teams

MyValueSolution — Fri, 17 Apr 2026 19:51:04 +0000

There’s a number floating around that should make every CFO pay attention. Visa — a payments company, not an AI lab — is now consuming 1.9 trillion AI tokens per month. That’s double what they were using just one month earlier. Eighty-nine per cent of their employees are active AI users, and nearly half qualify as power users, averaging 25 prompts a day.

Now zoom out. If a payments company is scaling AI this aggressively, what’s happening across financial services, fintech, and the thousands of companies building on open banking infrastructure? The answer: AI spend is exploding — and most organisations have no framework for controlling it. They have cloud budgets. They have procurement processes for SaaS tools. But AI consumption doesn’t fit neatly into either category. It’s usage-based, unpredictable, and growing faster than anyone forecasted.

This article is for the people who are staring at a cloud invoice that’s 40% higher than last quarter and trying to figure out what happened — and more importantly, what to do about it. Not the theoretical “AI governance framework” stuff. The practical stuff. What actually works.

The Problem: AI Costs Don’t Behave Like Other Software Costs

When you buy a SaaS product, you know what it costs. Ten seats at $50 per month. Done. Predictable. AI doesn’t work this way. API-based AI services charge per token — per input and per output. The cost of a single API call depends on which model you use, how long the prompt is, and how long the response is. A simple classification task might cost fractions of a cent. A complex document analysis task using a frontier model could cost several dollars per request.

This means your AI bill is a function of adoption — and adoption is the one thing your organisation is actively trying to increase. Every time a team builds a new automation, connects a new workflow, or rolls out an AI feature to customers, consumption goes up. The more successful your AI strategy is, the higher the bill gets. That’s the paradox finance teams are wrestling with right now.

The AI Spend Reality Check

Visa: 1.9 trillion tokens/month. Meta: 60 trillion tokens/month internally. A single developer at AI coding startup Cognition crossed 1 trillion tokens in cumulative usage. These are early indicators of where enterprise AI consumption is heading. If your company is using AI seriously, your cloud and API bills are going to grow — the question is whether that growth is managed or chaotic.

Seven Practical Ways to Reduce Your AI Spend

These aren’t hypothetical. These are the levers that actually move the number.

1. Use the Right Model for the Right Task

This is the single biggest cost lever most companies ignore. Not every task requires a frontier model. Classifying a support ticket? A smaller, cheaper model handles that just fine. Summarising a short email? Same. Generating a complex financial analysis from raw data? That’s where you want the expensive model. The mistake most teams make is defaulting everything to the most powerful (and most expensive) model because it was the first one they integrated. Build a routing layer that sends simple tasks to cheap models and complex tasks to premium models. The cost difference can be 10–50x per request.

2. Cache Everything You Can

If the same prompt produces the same output, you’re paying twice for the same answer. This is surprisingly common in production AI systems — especially in customer support, document processing, and data enrichment pipelines where the same types of queries repeat frequently. Implement a semantic caching layer that stores responses and serves them again when a sufficiently similar prompt arrives. Some teams report 20–40% cost reductions from caching alone.

3. Optimise Your Prompts

Longer prompts cost more — both on the input side and typically on the output side too, since verbose prompts tend to produce verbose responses. Review your production prompts and strip out unnecessary context, redundant instructions, and padding. A well-engineered prompt that’s 40% shorter produces the same quality output at 40% lower cost. This sounds trivial but across millions of API calls per month, the savings compound fast.

4. Use Batch Processing Where Latency Allows

Most AI providers offer batch APIs at significantly reduced rates — sometimes 50% cheaper than real-time inference. If your workload doesn’t need an instant response (think: overnight document processing, daily report generation, weekly data classification), batch it. The results are identical; you just wait a few hours instead of a few seconds. For back-office financial operations, this is often a trivial change that cuts the bill in half for those specific workloads.

5. Set Hard Spend Limits and Alerts

This sounds obvious but a shocking number of organisations don’t do it. Every major AI and cloud provider allows you to set monthly spend caps and usage alerts. Set them. Set them per team, per project, and per environment. A runaway loop in a development environment or an unanticipated spike in production traffic can burn through thousands of dollars in hours. The alert doesn’t reduce your spend — but it stops a manageable bill from becoming an emergency.

6. Audit Your Unused Credits and Commitments

Here’s something that doesn’t get talked about enough: many companies are sitting on cloud and AI credits they’re never going to use. Startup grants from Google Cloud, Azure, or AWS. Enterprise commitments that were over-provisioned. Promotional credits from provider partnerships. These credits expire — and when they do, the value is gone. If your organisation has unused Google Cloud capacity, you can sell Google Cloud credits through brokers who connect you with buyers looking for discounted capacity. It’s a straightforward way to recover cash from credits you’ll never consume instead of watching them expire on your balance sheet.

7. Negotiate — Or Find Alternative Supply

If your monthly AI spend exceeds $5,000 with any single provider, you almost certainly have room to negotiate. Contact your account representative and ask about volume pricing, committed use discounts, or custom rate cards. Most providers have these programmes but won’t proactively offer them unless you ask. And if your current provider won’t negotiate, explore alternatives — the model landscape is increasingly competitive and switching costs are lower than most teams assume.

Where the Money Actually Goes: AI Spend by Category

Most organisations don’t have clear visibility into what’s driving their AI costs. Here’s a typical breakdown for a mid-sized company running production AI workloads:

Cost Category	Typical Share	Primary Cost Driver
API Inference (Production)	35–45%	Real-time model calls in customer-facing features
Cloud Compute (Training & Fine-tuning)	15–25%	GPU hours for model customisation and experimentation
Data Processing & Embeddings	10–15%	Vector databases, RAG pipelines, search indexing
Internal Tools & Copilots	10–20%	Employee-facing AI assistants, code generation, analysis
Development & Testing	5–15%	Non-production model calls during development

The insight here is that production inference is usually the largest line item — but internal tools and development environments are often the fastest-growing categories, because they scale with employee adoption rather than customer demand. Visa’s numbers bear this out: their 44% power-user rate means almost half the company is generating significant token consumption every working day.

The Uncomfortable Truth: AI Spend Is a Feature, Not a Bug

Let’s be honest about something. If your AI spend is going up, it probably means your organisation is doing the right thing. The companies that will struggle in two years aren’t the ones spending too much on AI — they’re the ones spending too little. Visa isn’t tracking token consumption because they want to cut it. They’re tracking it because they want to reward the teams that are using it most effectively.

The goal isn’t to minimise AI costs in absolute terms. The goal is to maximise the value you get per dollar spent. That means eliminating waste (duplicate calls, oversized models for simple tasks, uncached repeated queries) while continuing to invest in the AI capabilities that actually move your business forward.

For finance and operations leaders in financial services and fintech, this is a familiar challenge dressed in new clothing. You’ve been optimising infrastructure spend for years — cloud migration, API gateway costs, payment processing fees. AI is just the newest line item. The tools are different but the discipline is the same: measure it, allocate it, and relentlessly cut the waste while protecting the investments that generate returns.

The Bottom Line

Your AI bill is going to keep growing. The question is whether you’re paying full price for everything or being smart about it. Route tasks to the right models, cache repeated queries, batch what you can, and don’t let cloud credits expire unused. The companies that treat AI spend as a strategic cost centre — not an uncontrolled line item — will have a structural advantage over those that don’t.

Frequently Asked Questions: AI Spend Management

Why is my AI spend increasing so fast?

AI costs are usage-based, not subscription-based. Every API call, every model inference, every embedding query costs money. As more teams adopt AI tools and more workflows get automated, consumption grows proportionally. Unlike a SaaS seat licence, there’s no natural ceiling — which is why active cost management is essential from the start.

What’s the fastest way to reduce AI API costs?

Route simple tasks to smaller, cheaper models instead of sending everything to the most expensive frontier model. This single change typically reduces costs by 30–60% with no measurable impact on output quality for routine tasks like classification, summarisation, and data extraction.

How much do different AI models actually cost?

Costs vary enormously by provider and model tier. Small, fast models can cost $0.10–0.25 per million input tokens. Large frontier models can cost $10–75 per million input tokens. That’s a 100x cost difference — which is why model selection is the most important cost lever available to engineering teams.

What is semantic caching and how does it save money?

Semantic caching stores the responses to AI queries and serves them again when a sufficiently similar prompt is received, avoiding a duplicate API call. Unlike exact-match caching, semantic caching uses embeddings to identify prompts that are similar in meaning even if the wording differs. Teams running high-volume, repetitive AI workloads typically see 20–40% cost reductions from caching.

Should I use batch APIs instead of real-time APIs?

Yes, wherever latency permits. Batch APIs are typically 50% cheaper than real-time inference. Any workload that doesn’t need an instant response — overnight processing, daily reports, weekly data classification, bulk document analysis — should be batched. The output quality is identical; you simply wait hours instead of seconds.

How do I know if I have unused cloud or AI credits?

Check your billing dashboard on each cloud and AI provider you use — Azure, AWS, Google Cloud, OpenAI, Anthropic. Look for prepaid credit balances, startup grants, promotional credits, or committed use allocations that are under-utilised. Many companies have credits they’ve forgotten about or that were provisioned by a previous team. These credits expire if unused.

Can I sell unused cloud credits instead of letting them expire?

Yes. There are brokers that connect sellers with buyers who want discounted cloud and AI credits. The process is confidential and the credits are transferred through legitimate mechanisms. If your organisation has credits that will expire before you can use them, selling them recovers real cash value instead of letting the credits disappear.

How much should my company be spending on AI?

There’s no universal benchmark. It depends on your industry, the maturity of your AI adoption, and the specific workloads you’re running. What matters more than the absolute number is the ratio of AI spend to value generated. Track the cost per automated task, cost per model inference, and cost per business outcome — then optimise those ratios rather than targeting an arbitrary budget number.

Is it worth negotiating with AI providers for better rates?

Absolutely. If you’re spending more than $5,000 per month with any single provider, you likely qualify for volume discounts, committed use pricing, or custom rate cards. Most providers have these programmes but won’t offer them unless you ask. The competitive landscape between providers also gives you leverage — if one won’t negotiate, another probably will.

What should a finance team track to manage AI costs?

At minimum: total AI spend by provider, spend by team or department, cost per API call by model, token consumption trends over time, and the ratio of production vs development/testing spend. Set monthly budgets, configure spend alerts, and review costs weekly during periods of rapid adoption. Treat it with the same rigour you apply to cloud infrastructure or payment processing costs.

The post How Do I Actually Lower My AI Spend? A Practical Guide for Finance Teams appeared first on My Value Solutions.

Solaris Cuts 20% of Workforce to Become Europe’s First AI-Native Bank — What It Means for BaaS

MyValueSolution — Tue, 31 Mar 2026 09:30:16 +0000

German embedded finance platform Solaris is eliminating approximately 80 positions — roughly 20% of its 400-person workforce — as part of what the company describes as a comprehensive strategic repositioning. The goal, according to new CEO Steffen Jentsch, is to transform Solaris from a Banking-as-a-Service provider into what the company is calling “Europe’s first AI-native bank.”

The restructuring follows a turbulent period for the Berlin-based fintech. Over the past three years, Solaris has navigated regulatory scrutiny from BaFin, multiple rounds of leadership changes, the failed Contis acquisition, and a rescue funding round in late 2024 that saw Japanese financial conglomerate SBI Group take a majority stake. The latest move under Jentsch — who took over as CEO just three months ago — represents the most definitive strategic shift in the company’s ten-year history.

What Solaris Is Actually Announcing

Beneath the headline, the repositioning has three concrete components. First, Solaris is developing an operating model in which AI agents handle routine operational processes — transaction monitoring, compliance workflows, customer onboarding — while human staff retain responsibility for control, governance, and regulatory oversight. Second, the company plans to restructure its platform around standardised, reusable modules rather than bespoke implementations for individual partners. Third, Solaris intends to expand beyond its traditional neobank and fintech client base toward larger enterprise and institutional partners.

The timing is deliberate. Jentsch has framed the pivot around two pieces of European regulation: the EU AI Act, which establishes the legal framework for deploying AI in high-risk sectors including financial services, and DORA (Digital Operational Resilience Act), which sets operational resilience standards for ICT systems in finance. The argument is that Europe is creating a regulatory environment that makes AI-native banking not only possible but competitively advantageous — and that Solaris, with its German banking licence and API-first architecture, is uniquely positioned to build within that framework.

“Ten years ago, Solaris was one of the first companies in Europe to prove that cloud-based banking via APIs works. Today, we are taking the next logical step.”

— Steffen Jentsch, CEO, Solaris

The Context That Matters: A Decade of Pivots

To understand the significance of this announcement, you need to understand how many times Solaris has already reinvented itself. The company was founded in 2015 as Solarisbank, incubated by the Berlin-based fintech company builder Finleap. It obtained its German banking licence in 2016 — remarkably fast by European standards — and built its early business by enabling neobanks and consumer fintechs like Vivid Money, Tomorrow, and Bitpanda to offer banking products through its API-driven platform without needing their own licences.

The company grew rapidly through the fintech funding boom, raising a total of approximately $736 million across multiple rounds and reaching a valuation of around €1.5 billion. It expanded into digital asset custody, identity verification, and crypto services. It won the Samsung Pay partnership in Germany. It acquired UK-based EMI Contis to expand into the UK market. Along the way, it dropped “bank” from its name in 2022, rebranding to simply Solaris — a signal that it saw itself as a technology company first.

But the growth-first narrative came at a steep cost. The Contis acquisition brought compliance failures, mass customer offboarding, and a lawsuit against Binance over unpaid fees. BaFin intervened with requirements to improve risk management and business organisation. In 2022, Solaris reported revenue of €130 million but losses of €56 million. In June 2025, BaFin fined the company €500,000 for repeatedly exceeding large exposure limits between 2022 and 2024. By the time SBI Group stepped in with €140 million in Series G funding in early 2025 — taking a 70% majority stake — Solaris was a company that had proven the BaaS model could work at scale but had failed to make it work profitably or compliantly.

What “AI-Native Banking” Actually Means — and What It Does Not

The phrase “AI-native bank” is doing a lot of heavy lifting in this announcement. Solaris is describing a model where AI agents handle operational processes — KYC verification, transaction monitoring, suspicious activity flagging, customer service interactions, regulatory reporting — while humans maintain oversight and governance. This is a meaningful architectural shift: instead of using AI to augment existing workflows, the company is proposing to build workflows around AI from the ground up, with human review as the exception rather than the rule.

The potential upside is clear. BaaS platforms carry enormous operational overhead. Each partner integration requires onboarding, compliance configuration, transaction monitoring calibration, and ongoing regulatory management. If AI can handle the repetitive, rules-based components of these workflows reliably, the unit economics of serving each partner improve dramatically. A platform that processes the same volume of transactions with half the operational staff would be fundamentally more competitive — and potentially more accurate, given that AI-driven compliance monitoring can operate continuously without the fatigue and inconsistency of manual processes.

But there is an obvious tension. Solaris’s most persistent problems over the past three years have been compliance failures — exactly the domain where the company now proposes to deploy AI most aggressively. BaFin fined the company for exceeding exposure limits. Its Contis subsidiary was fined for AML and information security failures. Regulators ordered improvements to business organisation and risk management. Deploying AI agents to handle compliance processes at a company with this regulatory track record will require extraordinary trust from supervisors — trust that Solaris has not yet earned.

The Klarna Precedent: Why “AI-First” Does Not Always Mean “AI-Only”

The European fintech sector already has a cautionary example of aggressive AI-driven workforce replacement. Swedish buy-now-pay-later giant Klarna replaced approximately 700 customer service roles with AI between 2022 and 2025, cutting its total headcount from around 7,400 to 3,000. The result was widespread customer complaints about declining service quality. Klarna’s CEO later acknowledged that the emphasis on cost savings had compromised the experience, and the company began rehiring human agents for complex cases — effectively abandoning the “full automation” model in favour of a hybrid approach.

The Klarna comparison is imperfect — Klarna’s AI deployment was primarily customer-facing, while Solaris is proposing AI-driven automation of back-office and compliance functions. But the underlying lesson is the same: the gap between what AI can handle in a controlled demonstration and what it can handle at production scale, with real edge cases, regulatory consequences, and reputational stakes, is often wider than executives expect. For a regulated banking infrastructure provider, the margin for error is even smaller than it is for a consumer lending platform.

The SBI Factor: Why the Majority Shareholder Matters

The involvement of SBI Group is the most underreported element of this story. SBI Holdings, led by chairman and CEO Yoshitaka Kitao, is one of Japan’s largest financial conglomerates with deep interests in digital finance, blockchain infrastructure, and cross-border payments. SBI’s acquisition of a 70% majority stake in Solaris was not a passive financial investment — it was a strategic acquisition designed to give SBI a licensed banking infrastructure platform in Europe.

Kitao’s public statement is instructive: he described the AI-native bank transformation as the mechanism through which SBI can “realise our vision of a digital financial infrastructure in Europe.” This positions Solaris not as an independent European fintech but as the European arm of SBI’s global digital finance ambitions. For Solaris’s existing BaaS partners — companies that chose Solaris specifically because it was an independent European platform with a German banking licence — this ownership shift raises questions about strategic alignment, data governance, and long-term platform direction that the AI narrative does not address.

What This Means for the European BaaS Market

Solaris’s repositioning has implications that extend well beyond the company itself. The European BaaS sector has consolidated significantly over the past two years. Railsr merged with Equals Money. Multiple smaller providers have exited or been absorbed. The market is increasingly divided between a handful of scaled platforms — Solaris, ClearBank, Banking Circle, ConnectPay — and a long tail of smaller, regionally focused providers.

If Solaris succeeds in dramatically reducing the operational cost of partner management through AI automation, it will set a new benchmark for what BaaS economics need to look like. Competitors that remain dependent on large compliance and operations teams will face a structural cost disadvantage. This could accelerate consolidation in a market that already has too many providers chasing too few profitable partnerships.

Conversely, if the AI-native approach encounters the same operational and regulatory difficulties that have plagued Solaris’s previous strategy shifts, it will reinforce the market’s growing scepticism about whether BaaS can be a sustainable business model at all. The sector has already lost significant investor confidence. Another high-profile stumble from its most prominent European player would make fundraising even more difficult for the companies still trying to scale.

The Bottom Line

Solaris’s transformation is equal parts ambitious and desperate. The company has exhausted most of its other strategic options. Growth-at-all-costs failed. The Contis acquisition failed. Diversification into crypto and identity failed. What remains is a German banking licence, an API platform, a Japanese majority shareholder with deep pockets and European ambitions, and a thesis that AI can fix the unit economics that a decade of venture-funded experimentation could not.

The thesis might be right. AI-driven compliance automation, intelligent transaction monitoring, and automated partner onboarding could genuinely transform the economics of banking infrastructure. But the execution risk is enormous — particularly for a company whose regulatory track record gives supervisors every reason to scrutinise rather than trust. The next twelve months will determine whether “AI-native banking” is a genuine paradigm shift or a rebranding exercise applied to the same structural challenges that have defined the BaaS sector’s difficult adolescence.

~80

Roles Eliminated

$736M

Total Raised

70%

SBI Ownership Stake

2015

Year Founded

Frequently Asked Questions

Solaris, AI-Native Banking, and the BaaS Market

What is Solaris and what does it do?

Solaris is a Berlin-based Banking-as-a-Service (BaaS) provider that holds a full German banking licence. It enables other companies — fintechs, retailers, software platforms — to offer banking products like accounts, cards, payments, and lending through its API-driven platform, without those companies needing to obtain their own banking licence. Founded in 2015, Solaris has served partners including Samsung, American Express, Bitpanda, and Vivid Money.

What does “AI-native bank” mean?

In Solaris’s usage, “AI-native” describes a banking platform where artificial intelligence agents handle the majority of operational processes — compliance checks, transaction monitoring, customer onboarding, regulatory reporting — as the default mode of operation, rather than as an add-on to existing human workflows. Humans remain responsible for oversight, governance, and exception handling, but AI is the primary processing layer. This is distinct from “AI-enhanced” banking, where AI tools assist human operators, and represents a more fundamental architectural commitment to automation.

Who is SBI Group and why are they involved?

SBI Holdings is a major Japanese financial services group with interests spanning online brokerage, asset management, insurance, digital assets, and fintech investments globally. SBI acquired a 70% majority stake in Solaris through a €140 million Series G round in early 2025, effectively taking control of the company. SBI views Solaris as its entry point into European banking infrastructure, using the company’s German banking licence and API platform as a foundation for its broader digital finance ambitions in the region.

What regulatory challenges has Solaris faced?

Solaris has faced sustained regulatory pressure over the past several years. Germany’s financial regulator BaFin ordered improvements to the company’s business organisation and risk management processes. In June 2025, BaFin fined Solaris €500,000 for repeatedly exceeding large exposure limits set by the European Capital Requirements Regulation between January 2022 and March 2024. Separately, its UK subsidiary Contis was fined €840,000 by the Bank of Lithuania for AML and information security failures. These compliance issues contributed to the leadership instability and strategic uncertainty that preceded SBI’s majority investment.

How does this affect Solaris’s existing BaaS partners?

Existing partners face both opportunity and uncertainty. The shift to standardised, reusable platform modules could reduce integration complexity and lower costs. However, the transition from bespoke partner configurations to a modular architecture may require migrations that disrupt existing integrations. The broader strategic shift — from serving primarily fintech startups toward enterprise and institutional clients — may also change the platform’s development priorities in ways that do not align with smaller partners’ needs. Partners relying on Solaris for compliance infrastructure will need clarity on how AI-driven compliance processes will maintain the accuracy and regulatory standing their own businesses depend on.

What role do the EU AI Act and DORA play in this strategy?

Solaris’s CEO has explicitly positioned the AI-native strategy within the context of two European regulations. The EU AI Act, which came into force in August 2024, establishes a risk-based framework for AI deployment in high-risk sectors, including financial services — creating legal clarity around what is and is not permissible. DORA (Digital Operational Resilience Act), which became applicable in January 2025, sets operational resilience standards for financial institutions’ ICT systems. Together, Solaris argues, these regulations create a compliance-friendly environment for AI in banking, because the rules of the road are now defined rather than ambiguous. Whether this regulatory framework is sufficient to satisfy supervisors overseeing a company with Solaris’s compliance history remains to be seen.

MyValue Solutions is an independent publication. We are not affiliated with Solaris, SBI Group, or any company mentioned in this article. This analysis represents our editorial assessment based on publicly available information and should not be construed as investment or financial advice.

The post Solaris Cuts 20% of Workforce to Become Europe’s First AI-Native Bank — What It Means for BaaS appeared first on My Value Solutions.

Europe’s First AI-Initiated Payment Just Ran on Live Banking Rails. Here’s Why That Matters.

MyValueSolution — Sun, 04 Jan 2026 09:45:16 +0000

Banco Santander and Mastercard have completed what they describe as Europe’s first live, end-to-end payment executed by an artificial intelligence agent. The transaction was not conducted in a sandbox or test environment — it ran on Santander’s live payments infrastructure, using real banking rails and real money, marking a significant milestone in the transition of agentic commerce from concept to production.

The pilot used Mastercard Agent Pay, an agentic payments protocol that enables AI agents to initiate and complete transactions on behalf of consumers within predefined permissions. PayOS handled the end-to-end orchestration. The result: an AI agent discovered a product, selected it, initiated payment, and completed the transaction — all without direct human intervention at the point of sale, while maintaining the security, governance, and consumer protection standards required under a regulated banking framework.

The announcement is significant not because an AI agent made a payment — that has been demonstrated in controlled environments for months — but because it happened on live banking infrastructure at one of the world’s largest financial institutions. With approximately €1.84 trillion in total assets, Santander’s involvement signals that agentic commerce is moving from fintech experimentation into mainstream banking.

What Actually Happened in the Pilot

The technical architecture of the transaction involves three distinct layers. Mastercard Agent Pay provides the payment protocol — a framework that registers and verifies AI agents, issues tokenised payment credentials, and ensures that every agent-initiated transaction operates within permissions defined by the consumer. The protocol builds on Mastercard’s existing tokenisation technology, the same infrastructure that secures mobile contactless payments and card-on-file transactions globally, extended to support autonomous AI agents as a new transaction initiator.

Santander contributed the live banking infrastructure — the actual payment rails, account systems, and regulatory compliance layer that processed the transaction. This is the element that distinguishes the pilot from previous agentic commerce demonstrations: the payment was not simulated or routed through a testing environment but processed through the same systems that handle Santander’s day-to-day transaction volume across Europe.

PayOS served as the orchestration layer, connecting the AI agent to the merchant, the payment protocol, and the banking infrastructure. This middleware function is critical because agentic commerce requires coordination between systems that were not originally designed to interact — conversational AI platforms, payment networks, merchant checkout systems, and bank processing engines all need to communicate in real time for an agent-initiated transaction to complete successfully.

The Mastercard Agent Pay Framework

Mastercard launched its Agentic Payments Programme in April 2025, introducing a framework designed to bring the same trust and security standards that govern traditional card payments into the emerging world of AI-initiated transactions. The core innovation is Mastercard Agentic Tokens — dynamic digital credentials that allow AI agents to transact on behalf of consumers while maintaining full authentication, permission boundaries, and transaction visibility.

The system works by requiring AI agents to be registered and verified before they can initiate payments. Each agent receives tokenised credentials tied to a specific consumer’s account, with predefined spending limits, merchant categories, and transaction types that the consumer has explicitly authorised. Every transaction initiated by an agent is traceable — the consumer, the issuing bank, and the merchant can all identify which agent acted, on whose behalf, and within what parameters.

Mastercard has been scaling the ecosystem rapidly. Citi and US Bank cardholders were the first to be enabled for Agent Pay-powered transactions, with all US Mastercard cardholders expected to be enabled by the 2025 holiday season and global rollout following shortly after. The network has integrated Agent Pay with PayPal’s wallet, giving hundreds of millions of consumers access to agentic commerce. It has also partnered with Stripe, Google, and Ant International’s Antom to make secure agentic transactions accessible to digital merchants globally. Live agentic transactions have now been completed in the US, Australia, New Zealand, Singapore, Malaysia, India, and South Korea — with the Santander pilot adding Europe to the list.

Why This Matters for Open Banking and Payment Infrastructure

The Santander-Mastercard pilot sits at the intersection of two major infrastructure trends that this publication covers closely: the evolution of open banking APIs and the emergence of AI agents as autonomous participants in financial transactions.

Open banking, as defined by PSD2 and the forthcoming PSD3/PSR framework, established the principle that third-party providers can access bank account data and initiate payments on behalf of consumers through standardised APIs. Agentic commerce extends this principle by adding a new category of third-party actor: the AI agent. Instead of a human consumer clicking through an online checkout, or a payment initiation service provider triggering a bank transfer, an AI agent autonomously discovers products, selects options, initiates payment, and completes the transaction — all within permissions that the consumer has defined in advance.

This raises a set of infrastructure questions that the payments industry is only beginning to address. How should AI agents be authenticated? How do existing Strong Customer Authentication (SCA) requirements apply when the transaction initiator is software rather than a human? How should consent be managed when an agent operates autonomously over time rather than receiving explicit approval for each individual transaction? How do merchants distinguish legitimate AI agents from automated bots attempting fraud?

Mastercard’s approach — registering agents, issuing purpose-specific tokens, and maintaining full transaction traceability — represents one answer to these questions. Visa has taken a parallel approach with its Trusted Agent Protocol, developed in collaboration with Cloudflare, which establishes a framework for secure communication between AI agents and merchants. The fact that both major card networks are investing heavily in agentic payment infrastructure suggests that the industry expects AI-initiated transactions to become a significant share of commerce volume within the next few years.

Santander’s AI Strategy: From ChatGPT Enterprise to Agentic Payments

Santander’s participation in the pilot reflects a broader AI strategy that has been building momentum over the past year. The bank’s AI initiatives are led by chief data and AI officer Ricardo Martín Manjón, who was hired from BBVA in March 2025. Under Manjón’s leadership, Santander formed a strategic partnership with OpenAI in August 2025, connecting up to 30,000 bank employees with ChatGPT Enterprise in what was described as one of the fastest deployments of its kind in the European banking sector.

The progression from internal AI deployment to live agentic payments is notable. Most banks that have adopted AI tools have focused on internal use cases — document processing, customer service augmentation, code generation, data analysis. Santander’s agentic payment pilot moves the AI deployment perimeter outward, into customer-facing transaction infrastructure where the stakes are materially higher. A chatbot that generates an imperfect internal summary is an inconvenience. An AI agent that executes an incorrect payment on live banking rails is a regulatory and reputational event.

The bank has confirmed it will move into extended testing and scaling, exploring additional use cases and partnerships while maintaining regulatory alignment. This measured language is significant — it signals that the pilot was a proof of concept, not a commercial launch, and that considerable work remains before agentic payments become a standard feature of Santander’s customer-facing product suite.

The Competitive Landscape: A Race to Define Agentic Commerce Standards

The Santander-Mastercard pilot does not exist in isolation. The past twelve months have seen an acceleration of agentic commerce activity across the payments industry. Visa launched its Trusted Agent Protocol in October 2025 and reported hundreds of secure agent-initiated transactions by December. DBS Bank in Singapore became the first issuer to pilot Visa Intelligent Commerce for everyday payments in February 2026. Mastercard showcased the first fully authenticated agentic commerce transaction in India using cards issued by Axis Bank and RBL Bank. South Korea’s first live agentic transaction was completed in March 2026 through the Mastercard Agent Pay framework.

The underlying competition is not just between banks or between card networks — it is a contest to define the standards and protocols that will govern how AI agents interact with the payments ecosystem. Mastercard is working with Google’s Universal Commerce Protocol, OpenAI’s Agentic Commerce Protocol, and other interoperability frameworks. Visa is building its own protocol stack. Meanwhile, technology companies like Google, OpenAI, and Microsoft are developing their own agent-to-merchant and agent-to-agent communication standards.

Whoever establishes the dominant protocol framework for agentic payments will hold a position of considerable influence over the next generation of digital commerce. The card networks are leveraging their existing trust infrastructure — tokenisation, fraud detection, dispute resolution — to argue that agentic payments should flow through established payment rails rather than through new, unregulated channels. This is a strategic play as much as a technical one: if AI agents transact primarily through card network infrastructure, Mastercard and Visa maintain their position as the central intermediaries of global commerce even as the transaction initiator shifts from human to machine.

What Comes Next

The Santander-Mastercard pilot is a milestone, but it is the beginning of a long integration process rather than a finished product. Several critical questions remain unresolved. Regulatory frameworks for agentic payments are still being developed — the European Banking Authority has not yet issued specific guidance on how existing payment services regulations apply to AI-initiated transactions. Consumer protection frameworks need to address scenarios where an AI agent makes a purchase that the consumer disputes. Liability allocation for agent errors — who is responsible when an AI agent exceeds its permissions or makes a suboptimal purchasing decision — remains an open legal question.

For banks, the opportunity is clear but the execution is complex. Agentic commerce represents a potential new distribution channel for payment products, a way to increase transaction volume, and an opportunity to deepen customer engagement through AI-assisted financial services. But it also requires banks to extend their infrastructure to support a new category of transaction initiator, to build or integrate agent management capabilities, and to adapt their compliance and risk frameworks to accommodate autonomous AI actors operating on live banking rails.

The financial institutions and payment networks that get this right will define how commerce works in the age of AI agents. Those that move too slowly risk being disintermediated by technology companies that build their own payment infrastructure. Those that move too fast risk the regulatory and reputational consequences of deploying AI in production financial systems before the governance frameworks are mature. The Santander-Mastercard pilot suggests that the smartest approach may be exactly what these two companies are demonstrating: controlled experiments on live infrastructure, conducted in close dialogue with regulators, with the ambition to scale but the discipline to test first.

Frequently Asked Questions

What is an agentic payment?

An agentic payment is a transaction initiated and completed by an AI agent on behalf of a consumer, without the consumer manually interacting with a checkout process. The AI agent discovers products or services, selects options based on the consumer’s preferences, initiates the payment using tokenised credentials, and completes the purchase — all within permissions that the consumer has defined in advance. The consumer retains control through predefined spending limits, merchant categories, and the ability to revoke agent permissions at any time.

What is Mastercard Agent Pay?

Mastercard Agent Pay is a payment protocol launched in April 2025 that enables AI agents to initiate and complete transactions securely on behalf of consumers. The system uses Agentic Tokens — dynamic digital credentials built on Mastercard’s existing tokenisation technology — to authenticate AI agents, enforce spending permissions, and maintain full transaction traceability. The protocol integrates with Microsoft Azure OpenAI Service and Microsoft Copilot Studio, and has been adopted by banks including Citi, US Bank, Westpac, and now Santander.

Why is the Santander pilot significant?

Previous agentic payment demonstrations were conducted in sandbox or test environments. The Santander pilot is significant because it was executed on the bank’s live payments infrastructure — real banking rails processing a real transaction. This demonstrates that agentic payments can operate within the security, governance, and regulatory compliance standards of a major European bank, moving the technology from proof-of-concept to production-grade viability.

How do agentic payments relate to open banking?

Open banking established the regulatory and technical framework for third-party providers to access bank account data and initiate payments through APIs with consumer consent. Agentic commerce extends this model by introducing AI agents as a new category of third-party actor that can autonomously initiate transactions. Both rely on the same foundational principles — consumer consent, secure API access, regulatory oversight — but agentic payments add complexity around authentication, autonomous permission management, and the governance of non-human transaction initiators.

What is the difference between Mastercard Agent Pay and Visa’s Trusted Agent Protocol?

Both are frameworks designed to enable AI agents to initiate secure payments, but they take different approaches. Mastercard Agent Pay focuses on extending its existing tokenisation infrastructure to issue agent-specific credentials, with integration points into Microsoft, PayPal, Stripe, and Google ecosystems. Visa’s Trusted Agent Protocol, developed with Cloudflare, establishes a framework for secure communication between AI agents and merchants, focusing on verifying agent identity and distinguishing legitimate agents from bots. Both networks are actively working with regulators and industry bodies to define interoperable standards for agentic commerce.

When will agentic payments be available to consumers?

Agentic payments are already available in limited form. All US Mastercard cardholders are expected to be enabled for Agent Pay by mid-2026, with global rollout following. Live agentic transactions have been completed in the US, Australia, New Zealand, Singapore, Malaysia, India, South Korea, and now Europe through the Santander pilot. However, widespread consumer adoption will depend on the maturation of AI agent platforms, merchant acceptance, regulatory clarity, and consumer willingness to delegate purchasing authority to AI systems. Most industry observers expect agentic commerce to become a meaningful share of transaction volume within three to five years.

MyValue Solutions is an independent publication. We are not affiliated with Santander, Mastercard, or any company mentioned in this article. This analysis represents our editorial assessment based on publicly available information and should not be construed as investment or financial advice.

The post Europe’s First AI-Initiated Payment Just Ran on Live Banking Rails. Here’s Why That Matters. appeared first on My Value Solutions.

Global Fintech Investment Rebounds to $116 Billion in 2025 — But Fewer Deals Signal a Market That Rewards Scale Over Ambition

MyValueSolution — Mon, 04 Aug 2025 11:16:56 +0000

Global fintech investment turned a corner in 2025. After three consecutive years of decline — from a peak of $168.4 billion in 2022 to a seven-year low of $95.5 billion in 2024 — total investment rebounded to $116 billion, according to KPMG’s latest Pulse of Fintech report, published in February 2026. The recovery was driven by larger deal sizes, a resurgence in exit activity, and surging investor interest in two dominant themes: digital assets and artificial intelligence.

But the headline number obscures a more nuanced reality. While capital deployment increased by 21% year-over-year, the total number of deals fell to 4,719 — an eight-year low and the fourth consecutive annual decline. Investors are not spreading their bets more widely; they are concentrating capital into fewer, larger, later-stage companies with proven business models and clear paths to profitability. The era of broad-based fintech funding optimism has not returned. What has returned is selective confidence in specific sectors and companies that have survived the downturn.

For anyone operating in open banking, financial data infrastructure, or payments technology, the report contains critical signals about where institutional capital is flowing, which fintech subsectors are gaining traction, and what the competitive landscape will look like over the next twelve months.

The Global Picture: More Money, Fewer Deals

The overall investment picture for 2025 tells a story of recovery in capital deployed but continued contraction in deal activity. The $116 billion total represents a meaningful increase from 2024’s $95.5 billion, but remains well below the $168.4 billion peak of 2022 and the $119.3 billion recorded in 2023. Venture capital accounted for the largest share at $56.7 billion across 3,765 deals, followed by M&A at $55.3 billion across 840 deals. Private equity growth funding declined from $5.5 billion to $4 billion.

Global Fintech Investment by Year (2022–2025)
Year	Total Investment	Deal Count	VC Investment	M&A Value	PE Growth
2022	$168.4B	8,314	$92.1B	$65.6B	$10.7B
2023	$119.3B	5,764	$51.1B	$58.6B	$9.6B
2024	$95.5B	5,533	$45.4B	$44.6B	$5.5B
2025	$116.0B	4,719	$56.7B	$55.3B	$4.0B
Source: KPMG Pulse of Fintech H2’25 (data provided by PitchBook, as of 31 December 2025)

The decline in deal count is particularly significant for early-stage fintech companies. Investors are increasingly channelling capital into late-stage rounds and proven platforms rather than seeding new entrants. Median pre-money valuations at the venture growth stage surged from $168.3 million in 2024 to $977.5 million in 2025 — a nearly six-fold increase that reflects the concentration of capital in a shrinking number of category-defining companies. For founders raising seed or Series A rounds, the funding environment remains materially more difficult than the headline investment figures suggest.

Regional Breakdown: Americas Dominates, EMEA Recovers, Asia-Pacific Struggles

The Americas accounted for more than half of global fintech investment in 2025, attracting $66.5 billion across 2,409 deals. The US alone represented $56.6 billion of this total, up from $42.4 billion in 2024. The American recovery was driven by a combination of large late-stage VC rounds, a reopening IPO market, and strong activity in the digital assets space following the passage of the GENIUS Act.

The EMEA region saw a modest but meaningful recovery, with investment rising to $29.2 billion from a 2024 low, despite deal volume falling to an eight-year low of 1,484. The UK retained its position as Europe’s dominant fintech hub with $10.9 billion in investment, followed by the Nordics with a strong $5.3 billion — of which Sweden accounted for $4.8 billion. The region saw growing regulatory momentum, with the European Parliament agreeing to both the Payment Services Regulation (PSR) and the Third Payment Services Directive (PSD3) during H2’25, providing greater clarity for open banking and payments companies operating across the EU.

Asia-Pacific was the weakest region, with investment falling to a ten-year low of $9.3 billion across just 763 deals. China’s fintech sector continued to contract amid economic challenges and regulatory tightening, while Australia experienced rightsizing. South Korea was a bright spot, with investment nearly doubling to $402 million, though much of that was concentrated in a single deal — fintech super app Toss’s $200 million raise.

Fintech Investment by Region (2024 vs 2025)
Region	2024 Investment	2024 Deals	2025 Investment	2025 Deals	YoY Change
Americas	$55.4B	2,627	$66.5B	2,409	+20.0%
EMEA	$26.5B	1,803	$29.2B	1,484	+10.2%
Asia-Pacific	$11.7B	1,028	$9.3B	763	−20.5%
Global Total	$95.5B	5,533	$116.0B	4,719	+21.5%
Source: KPMG Pulse of Fintech H2’25 (data provided by PitchBook, as of 31 December 2025)

Digital Assets: The Dominant Theme of 2025

The digital assets sector was the standout investment story of 2025. Total global investment in the space nearly doubled year-over-year, rising from $11.2 billion to $19.1 billion. While the total remained below the $32.2 billion record set in 2021, the current trajectory — driven by regulatory certainty rather than speculative frenzy — suggests a more sustainable growth path.

The catalyst was regulatory clarity. In the US, the passage of the GENIUS Act (Guiding and Establishing National Innovation for US Stablecoins) in H2’25 gave banks and institutional investors the legal framework they needed to participate in the digital assets ecosystem with confidence. In Europe, MiCA (Markets in Crypto-Assets) came into full force at the end of 2024, establishing a comprehensive regulatory regime across the EU. The UK announced plans for its own regulatory framework, expected to be in force by 2027.

Stablecoins attracted particularly intense interest. A consortium of major European banks — including ING, BNP Paribas, KBC, and UniCredit — announced the creation of Qivalis, a new entity mandated to launch a euro-pegged stablecoin by end of 2026. Separately, a group of global systemically important banks including Bank of America, Deutsche Bank, Goldman Sachs, UBS, and Citigroup announced plans to jointly explore issuing a stablecoin pegged to G7 currencies. Asset tokenisation — particularly of money market funds and real estate — also gained momentum, with BlackRock announcing plans to tokenise its top-performing ETFs and Fidelity launching a digital interest token fund.

The IPO market for digital asset companies reopened in a meaningful way during 2025, with blockchain lender and stablecoin issuer Figure raising $787.5 million and crypto exchange Gemini raising $425 million in Nasdaq listings during H2’25. Coinbase’s $2.9 billion acquisition of crypto derivatives exchange Deribit was the largest deal in the sector’s history.

AI-Focused Fintech: $16.8 Billion and Growing

AI-focused fintech companies attracted $16.8 billion in global investment in 2025, up from $12.1 billion in 2024, with deal volume rising from 1,183 to 1,334. Corporate investors were particularly active, driven by a focus on AI solutions that could deliver operational efficiencies and cost savings in areas like compliance automation, fraud detection, credit decisioning, and customer service.

However, KPMG’s analysis reveals an important distinction: the majority of corporate AI investment flowed to partnerships with large technology and AI companies rather than to fintech startups. Banks and financial institutions increasingly chose to work directly with established AI providers to develop internal capabilities rather than relying on third-party fintech vendors. This trend was particularly visible in the regtech space, where total investment actually declined from $6.8 billion to $4.9 billion even as banks accelerated their internal AI compliance programmes.

The implication for AI-focused fintech startups is clear: to attract meaningful investment, they will need to develop differentiated intellectual property that large institutions cannot easily replicate through partnerships with big tech. Generic AI wrappers around existing processes will not be sufficient. The winners will be companies that enable genuine business model transformation or deliver capabilities that are specific enough to financial services that general-purpose AI platforms cannot match.

Sector-by-Sector: Where the Money Went

The KPMG report breaks down investment across six fintech subsectors, revealing significant variation in momentum and investor sentiment.

Fintech Investment by Subsector (2024 vs 2025)
Subsector	2024 Investment	2024 Deals	2025 Investment	2025 Deals	YoY Change
Payments	$20.4B	655	$19.2B	542	−5.9%
Digital Assets	$11.2B	1,584	$19.1B	1,199	+70.5%
AI-Focused Fintech	$12.1B	1,183	$16.8B	1,334	+38.8%
Insurtech	$2.9B	298	$8.6B	291	+196.6%
Regtech	$6.8B	431	$4.9B	519	−27.9%
Wealthtech	$4.9B	58	$1.4B	57	−71.4%
Cybersecurity	$0.9B	93	$0.7B	72	−22.2%
Source: KPMG Pulse of Fintech H2’25 (data provided by PitchBook, as of 31 December 2025)

Payments investment was relatively flat at $19.2 billion, but the composition shifted significantly. Investors moved away from consumer-focused models toward B2B payments infrastructure, and emerging markets — particularly South America — attracted growing attention. Revolut’s $3 billion raise at a $75 billion valuation was the sector’s landmark deal, while B2B payments, real-time payments monetisation, and cross-border settlement infrastructure dominated the deal pipeline.

Insurtech investment tripled on the back of two outlier deals — the $2.6 billion acquisition of Next Insurance by Ergo and the $2.5 billion take-private of Sapiens International by Advent — but underlying deal volume remained soft. Wealthtech investment collapsed from $4.9 billion to $1.4 billion, reflecting both the absence of large outlier deals and a broader shift of investor attention toward AI and digital assets.

The regtech decline is particularly significant for the open banking ecosystem. Despite a growing regulatory burden — PSD3, DORA, the EU AI Act, MiCA — investment in standalone regtech companies fell as banks increasingly chose to build AI-driven compliance capabilities internally rather than buy them from third-party vendors. This is a structural shift that may permanently reshape the regtech market, accelerating consolidation and pushing smaller players toward early exits via acquisition.

The Exit Market Reopens

Perhaps the most encouraging signal in the entire report is the recovery in fintech exit activity. Global exit value more than doubled year-over-year, from $46.8 billion to $104.4 billion — the third highest level on record. The number of exits increased from 438 to 486. Critically, the US IPO market reopened for fintech companies in a meaningful way, with VC-backed fintech IPOs generating $63 billion in global exit value — the second highest annual total after the 2021 boom.

The reopening of exit markets matters for the entire fintech funding ecosystem. VC and PE investors need exits to generate returns and recycle capital into new investments. The absence of viable exits during 2022-2024 created a bottleneck that depressed both fundraising and deal activity. With exit paths reopening — particularly through IPOs and strategic M&A — the conditions are in place for a more robust investment cycle in 2026.

What This Means for Open Banking and Financial Infrastructure

Several trends in the KPMG report have direct implications for the open banking, payments, and financial data infrastructure space that MyValue Solutions covers.

The concentration of payments investment in B2B infrastructure and real-time settlement is consistent with the broader shift from consumer-facing fintech to backend plumbing. Companies building the middleware that connects banks, payment processors, and third-party providers — the exact infrastructure layer that open banking depends on — are attracting institutional capital at a level that suggests investors see long-term value in this segment.

The regulatory momentum around PSD3, PSR, and FIDA in Europe creates both opportunity and complexity for open banking platforms. The KPMG report notes that the European Parliament agreed to both the PSR and PSD3 during H2’25, which will harmonise payment services regulations and create a more competitive environment for third-party providers. For companies operating in this space, the regulatory direction is unambiguously positive — but implementation timelines and technical standard-setting will determine how quickly the benefits materialise.

The decline in regtech investment, paradoxically, may benefit open banking infrastructure providers that embed compliance capabilities directly into their platforms. As banks move away from point-solution regtech vendors toward integrated AI-driven compliance tools, the companies best positioned to capture this demand are those that offer compliance as a feature of a broader infrastructure offering — not as a standalone product.

Finally, the emergence of agentic commerce — AI agents that can initiate and complete financial transactions autonomously — connects the AI investment trend directly to the payments and open banking infrastructure layer. As KPMG notes, agentic commerce is expected to drive “channel and distribution disruption in the consumer domains, while also taking hold in payment financial market infrastructure in wholesale markets.” For open banking APIs and payment initiation services, the rise of AI agents as a new category of transaction initiator represents a significant expansion of the addressable market — and a new set of infrastructure requirements that do not yet exist at scale.

Looking Ahead: KPMG’s Top Predictions for H1 2026

KPMG identifies five key trends to watch in the first half of 2026. Digital assets and tokenisation will remain the dominant narrative, with stablecoins, tokenised deposits, and real-world asset tokenisation all accelerating. AI will continue to attract the largest share of overall investment, though corporates will prioritise partnerships with big tech over investments in fintech startups. Capital markets are expected to see significant disruption as startups targeting equity trading, debt capital markets, and private credit mature. Asset management — particularly in Asia-Pacific — is positioned for a technology-driven transformation, and the UAE is moving aggressively to position itself as a global hub for fintech and real-world asset tokenisation.

The overall tone of the report is cautiously optimistic. KPMG’s global fintech lead Anton Ruddenklau describes a market that is “finding its footing again” — one where macroeconomic and geopolitical risks remain real, but where the combination of stronger exit markets, regulatory clarity, and accelerating innovation provides a constructive foundation for sustained investment. For companies building in the open banking and financial infrastructure space, the message is clear: the funding environment is improving, but selectivity is the defining feature. The companies that will attract capital in 2026 are those with proven revenue models, differentiated technology, and a clear path to profitability — not those that are simply riding the AI or digital assets narrative without substance behind it.

Frequently Asked Questions

How much was invested in fintech globally in 2025?

According to KPMG’s Pulse of Fintech report, total global fintech investment reached $116 billion across 4,719 deals in 2025, up from $95.5 billion across 5,533 deals in 2024. The increase in investment despite a decline in deal count reflects investor concentration on larger, later-stage deals with companies that have established business models and clear paths to profitability.

Which fintech sectors attracted the most investment in 2025?

Payments remained the largest fintech subsector by investment volume at $19.2 billion, followed closely by digital assets at $19.1 billion — which nearly doubled year-over-year. AI-focused fintech companies attracted $16.8 billion, up from $12.1 billion in 2024. Insurtech saw a dramatic increase to $8.6 billion, though this was driven by two outlier deals rather than broad-based growth.

What drove the surge in digital assets investment?

Regulatory clarity was the primary catalyst. The passage of the GENIUS Act in the US provided a legal framework for stablecoins, while the EU’s MiCA regulation came into full effect. These developments gave banks and institutional investors the confidence to participate in the digital assets ecosystem. Stablecoins, asset tokenisation (particularly money market funds), and strong IPO activity from companies like Figure and Gemini all contributed to the investment surge.

Why did regtech investment decline despite growing regulation?

Regtech investment fell from $6.8 billion to $4.9 billion despite an expanding regulatory landscape. KPMG attributes this to banks and financial institutions increasingly building AI-driven compliance capabilities internally rather than purchasing from third-party regtech vendors. This structural shift, combined with the fragmented and niche nature of most regtech startups, is driving early exits through acquisition and challenging the long-term viability of standalone regtech business models.

What does the report say about fintech exits and IPOs?

Fintech exit activity recovered strongly in 2025, with global exit value more than doubling from $46.8 billion to $104.4 billion — the third highest level on record. The US IPO market reopened meaningfully for fintech companies, with VC-backed fintech IPOs generating $63 billion in global exit value. Digital asset companies were particularly active on the IPO front, and post-IPO performance was generally positive, which bodes well for the pipeline of companies considering public listings in 2026.

MyValue Solutions is an independent publication. We are not affiliated with KPMG, PitchBook, or any company mentioned in this article. All data cited is sourced from KPMG’s Pulse of Fintech H2’25 report (data provided by PitchBook, as of 31 December 2025). This analysis represents our editorial assessment and should not be construed as investment or financial advice.

The post Global Fintech Investment Rebounds to $116 Billion in 2025 — But Fewer Deals Signal a Market That Rewards Scale Over Ambition appeared first on My Value Solutions.