AI in Financial Services & Wealth Management

The shape of the evidence: near-universal adoption, early impact

Every major survey of the sector now reports majority or near-universal use of generative AI, and the trajectory is steep. NVIDIA’s fifth annual State of AI in Financial Services survey found that “more than half” of financial professionals — 52% — now use generative AI, up from 40% the prior year.² Among the more specialised population of wealth and asset managers, EY-Parthenon reports that 95% of firms have scaled adoption to multiple use cases and 78% have implemented three to five distinct use cases.¹⁰ Practitioner sentiment has flipped in parallel: in Advisor360°‘s survey of 300 U.S. advisors, 85% now call generative AI a “help” to their practice (up from 64% a year earlier), while the share calling it a “threat” to their livelihood fell from 21% to 8%.⁹

The harder finding is the gap between adoption and demonstrated value. EY’s same population reports only 27% seeing a “substantial impact on their firm over the past one-to-two years” — a wide expectation-versus-impact gap given that 95% have scaled.¹⁰ That is consistent with the cross-industry baseline: Stanford’s 2025 AI Index, drawing on McKinsey survey data, concludes that “most companies that report financial impacts from using AI within a business function estimate the benefits as being at low levels,” with the most common cost decrease under 10% and the most common revenue increase at or below 5%.¹² NVIDIA’s respondents are more bullish — “nearly 70%” report AI driving a revenue increase of 5% or more, and “more than 60%” report cutting annual costs by 5% or more² — but NVIDIA surveys a self-selected, AI-engaged population (executives, data scientists, developers), which plausibly skews the impact figures upward relative to the whole industry.

Where the money is: the value on paper, and the spend behind it

The most-cited estimate of the prize is McKinsey’s: across the global banking sector, generative AI “could add between $200 billion and $340 billion in value annually, or 2.8 to 4.7 percent of total industry revenues, largely through increased productivity” — equivalent to 9–15% of operating profits.⁸ Banking ranks among the highest-impact industries as a share of revenue because so much of its work is knowledge work: customer operations, marketing and sales, software engineering, and risk and legal review.⁸ The greatest absolute gains fall in the corporate and retail segments, at roughly $56 billion and $54 billion respectively.⁸ This sits inside McKinsey’s wider cross-economy headline of $2.6–4.4 trillion in annual value.⁸

The spend behind those estimates is concrete. Among EY’s surveyed wealth and asset managers, 75% are budgeting investments exceeding $11 million, and one-third plan significant resource increases within two years.¹⁰ Deloitte’s financial-services cut of its enterprise survey (roughly 540 FS leaders) shows the spend concentrating in a leading cohort: it splits respondents into “pioneers” (46%) and “followers” (54%), and on every dimension the pioneers are well ahead. Seventy-six percent of pioneers allocate 20% or more of their AI budget to generative AI, versus 46% of followers; 74% of pioneers estimate ROI exceeding 10%, versus 44% of followers; and 47% of pioneers say ROI exceeded expectations, against just 17% of followers.¹¹ The gap is a maturity gap, not a technology gap — followers most often cite missing adoption strategy, talent, and governance as the constraint.¹¹

The use-case map: advice, planning, service, compliance, fraud, operations

The applications cluster, and they are mostly internal-efficiency and customer-experience plays rather than autonomous advice. NVIDIA’s respondents rank trading and portfolio optimisation as the top generative-AI use case by ROI (25% of responses), followed by customer experience and engagement (21%).² The single largest shift the survey records is in customer-facing technology: chatbot and virtual-assistant use “surged from 25% to 60%,” and more than half of respondents now use generative AI for document processing and report generation.² Stanford’s function-level data points the same way — the most common cost savings come from service operations, supply-chain management, and software engineering, while the most common revenue gains come from marketing and sales and from service operations.¹²

Advice, planning, and client service

In wealth specifically, advisor sentiment data shows appetite concentrated in augmentation — administrative assistance, prospecting, and predictive analytics — rather than replacement of judgment; Advisor360°‘s president frames the thesis as “the true promise of AI isn’t in replacing human judgment — it’s in amplifying it.”⁹ EY observes the next step already underway: 78% of surveyed wealth and asset managers are exploring agentic AI to unlock deeper strategic advantage, even as the overwhelming majority keep humans firmly in the loop.¹⁰

Compliance, AML, and operations

On the compliance side, the picture is deliberately cautious. FINRA’s 2025 oversight report finds firms “proceeding cautiously,” generally exploring or implementing third-party vendor-supported generative-AI tools “to increase efficiency of internal functions” — summarising across multiple sources into one document, conducting analyses across disparate data sets, and helping employees retrieve relevant portions of policies and procedures.¹⁹ That same report flags the inverse risk in fraud and cyber: threat actors are exploiting generative AI for “fake content, polymorphic malware, and other malicious tools,” and treats AML and fraud as continuing supervisory focus areas.¹⁹ The OECD’s 49-jurisdiction survey similarly notes generative AI being used “as a tool to create sophisticated phishing messages” — the technology cuts both ways across fraud detection and fraud commission.²¹

Robo-advice: the history, and the limits

Robo-advice is the sector’s longest-running mass-market AI experiment, and its arc is instructive. The independent Robo Report puts the tracked sector at roughly $1.2 trillion in assets at year-end 2024, up from $1.089 trillion in 2023.⁷ But the headline conceals two things. First, concentration: the bulk of those assets sit with incumbents’ hybrid offerings — Vanguard’s combined advice assets of $365.1 billion and Edelman Financial Engines at $292.9 billion dwarf the digital-native pure-plays such as Betterment ($56.4B) and Wealthfront ($35.3B).⁷ Second, deceleration: client growth is slowing even as AUM rises — Betterment added clients at roughly 9%, Wealthfront at around 8%, and Acorns’ net new sign-ups slowed to roughly 4%.⁷

The economics of pure automation have proved hard. JPMorgan shut down its Automated Investing robo in Q2 2024, with the bank stating plainly that “the robo-investing business did not take off in the wealth industry as expected. It hasn’t scaled or become profitable for many, including us.”¹⁸ The same report catalogues a broader retreat — Betterment layoffs, BlackRock’s FutureAdvisor sold on, a Titan SEC fine — and quotes an analyst’s read that robos struggle to “demonstrate sustainable profitability,” with the sector pivoting toward hybrid models that use digital experiences to spark interest and then connect clients with human advisors.¹⁸

The academic critique explains why pure robo-advice hits a ceiling. A behavioural-finance review of 80 peer-reviewed sources identifies four structural limits: a service-relationship gap, because algorithms cannot build the affective trust central to financial relationships; algorithmic bias, because “algorithms designed by humans … cannot be completely free from human affect, cognition, or opinion”; market-risk persistence, because optimisation within constraints cannot eliminate systematic risk; and financial-literacy stagnation, because passive automation “does not actively engage users in the learning process” and may foster overconfidence.¹⁷ The hybrid pivot is, in effect, the industry conceding the first limit.

Regulation and oversight: existing rules, applied to new technology

The dominant regulatory posture in finance is technology-neutral: apply the rules already on the books rather than write AI-specific ones. The OECD’s survey of 49 jurisdictions finds that “the vast majority of respondent jurisdictions have introduced some form of policy that covers AI” in parts of finance, but “only a minority of financial regulators/supervisors … have issued specific” AI guidance; more than a dozen rely on non-binding principles, strategies, and white papers, and the overall approach is “risk-based and technology-neutral.”²¹ Binding, AI-specific or cross-sectoral legislation touching finance exists in a handful of places — the EU AI Act and laws in Brazil, Chile, and Colombia — but even the EU AI Act has explicit provisions covering only part of the financial sector.²¹

FINRA’s stance is the cleanest statement of the principle. Its Regulatory Notice 24-09 reminds members that existing rules apply to generative AI and large language models, and is explicit that the notice “does not create new legal or regulatory requirements or new interpretations of existing requirements.”²⁰ Rule 3110 (Supervision) still requires a reasonably designed supervisory system — and FINRA specifies that policies for AI in compliance review should address “technology governance, including model risk management, data privacy and integrity, reliability and accuracy of the AI model”; Rule 2210 (Communications) applies equally to human- and technology-generated content.²⁰ FINRA’s 2025 oversight report adds the supervisory worry in plainer language: firms’ use of generative AI is “outpacing the controls, documentation and supervisory frameworks needed to manage the technology’s risks.”¹⁹

The SEC: enforcement first, rulemaking withdrawn

The SEC’s most concrete action to date is enforcement, not rulemaking. On 18 March 2024 it brought its first “AI washing” cases against two investment advisers. Delphia (USA) Inc. had falsely claimed it used “machine learning” and “artificial intelligence” to analyse client data despite having admitted during a 2021 examination that no such algorithm existed; Global Predictions, Inc. had falsely marketed itself as the “first regulated AI financial advisor” offering “expert AI-driven forecasts.”⁴ The two settled for civil penalties of $225,000 and $175,000 respectively — $400,000 combined — under the Investment Advisers Act anti-fraud provisions and the Marketing Rule.⁴ The actionable lesson for the industry is narrow and clear: the SEC is policing claims about AI, regardless of the underlying technology.

The headline rulemaking effort, by contrast, has been abandoned. In August 2023 the SEC proposed rules on “Conflicts of Interest Associated with the Use of Predictive Data Analytics by Broker-Dealers and Investment Advisers” (File No. S7-12-23), which would have required firms to eliminate or neutralise conflicts arising from predictive-data-analytics and AI-driven tools used in investor interactions. On 12 June 2025 the SEC formally withdrew that proposal — one of fourteen rule proposals from the prior administration withdrawn at once.⁶ The practical status: any future rulemaking on predictive data analytics “must start anew with a new proposal and a fresh opportunity for public comment.”⁶

Risks specific to the sector: model risk, bias, opacity, privacy, concentration

The risks that regulators and standard-setters single out for finance are not generic AI risks; they are the ones amplified by leverage, interconnection, and fiduciary duty. The IMF’s FinTech Note on generative AI in finance lays out the canonical taxonomy: embedded bias, privacy concerns, outcome opaqueness, performance robustness, unique cyberthreats, and “the potential for creating new sources and transmission channels of systemic risks” — concluding that generative AI “could aggravate some of these risks and bring about new types of risks as well, including for financial sector stability.”¹³ The Financial Stability Board’s 2024 report converges on four: third-party dependency and concentration risk, correlated market behaviour (“herding”), expanded cyber attack surface, and model risk and data-governance failures.¹⁴

Bias and fair lending

The most rigorously measured sector-specific harm is lending discrimination, and the finding is genuinely two-sided. The NBER study of U.S. mortgage pricing found that lenders charge Latinx and Black borrowers 7.9 and 3.6 basis points more on purchase and refinance mortgages respectively — roughly $765 million per year in extra interest in aggregate.¹⁵ Algorithmic FinTech lenders also discriminate on price, but “40% less than face-to-face lenders,” and the authors find FinTechs do not discriminate in loan approval at all, even though an estimated 0.74–1.3 million minority applications were rejected between 2009 and 2015 due to discrimination overall.¹⁵ Algorithms inherit bias but, in this case, less of it than the humans they replaced — which is an argument for scrutiny of the data and the model, not for assuming either bias or fairness.

Explainability, fiduciary duty, and data privacy

Opacity collides directly with fiduciary obligation. FINRA’s insistence that AI governance address “reliability and accuracy of the AI model” is, in substance, an explainability requirement: a firm that cannot explain a recommendation cannot demonstrate it acted in the client’s interest.²⁰ Data privacy is the most-cited operational concern by practitioners — 77% of EY’s surveyed wealth and asset managers cite concerns around data privacy, accuracy, and external-data use, and the response is defensive: 87% rely on closed-source models on trusted platforms and 86% were surprised by the regulatory and compliance complexity.¹⁰ FINRA’s concrete guardrail is to add contractual language “that prohibits firm or customer sensitive information from being ingested into a third-party vendor’s open-source Gen AI tool.”¹⁹

Systemic concentration and the monoculture problem

The risk most distinctive to finance is systemic, and it traces to concentration. Then-SEC Chair Gary Gensler warned that institutions will concentrate on a small number of large base models, so that “the whole financial sector, indirectly, will be relying on those central nodes” — and if those nodes “have it wrong, the monoculture goes one way, well, then there’s a risk in society.”⁵ He framed this as largely beyond the SEC’s reach, since regulators are built around “entities and activities,” not foundational technologies, and called for diversity of models and data sources to avoid “a pretty fragile system.”⁵ The Roosevelt Institute sharpens the mechanism for agentic AI: when many agents rely on similar models and data they can react identically, potentially triggering “bank runs and flash crashes,” while dependence on a handful of providers means a single failure could produce “cascading effects throughout the financial system.”¹⁶ The Roosevelt analysis also flags a fiduciary trap unique to “agents”: despite the name, they “are not guaranteed to act in users’ interests” and may be designed to “preference the provider.”¹⁶

Reading the gap honestly

The consistent through-line across every source — vendor, consultancy, regulator, and academic — is a wide and deliberate gap between adoption and realised impact. Firms have adopted fast (52% of professionals, 95% of surveyed managers) and spent real money,²¹⁰ but the substantial-impact figure is 27% and the typical reported benefit is small.¹⁰¹² The binding constraint is not capability; it is governance, talent, and regulatory uncertainty — which is exactly why 86% of surveyed managers say regulatory and compliance complexity caught them by surprise, and 88% of asset managers and 84% of wealth managers name regulatory compliance their greatest hurdle.¹⁰ Robo-advice is the cautionary precedent: a decade of automation produced a trillion dollars in assets but not pure-play profitability, and the market resolved toward hybrids that keep a human in the relationship.⁷¹⁸

The regulatory and risk picture reinforces the cautious read. Oversight is currently delivered through existing technology-neutral rules and selective enforcement rather than a comprehensive AI regime,²⁰²¹ the one prominent AI-specific U.S. rulemaking was withdrawn,⁶ and the risks that distinguish finance — model opacity against fiduciary duty, measured bias, data privacy, and AI-driven concentration — are precisely the ones that slow responsible deployment.¹³¹⁴⁵ The defensible industry-level conclusion is that the technology is past proof-of-concept and short of transformation: real, near-universal, and so far additive at the margin rather than structurally reshaping the economics of advice, planning, or risk.

References