2026 Outlook AI Transformation of the IT Function in Financial Services

From productivity tools
to operating-model
reinvention.

Generative and agentic AI now sit on top of an IT function that regulators already supervise tightly. The question is no longer whether AI can accelerate a developer or summarize a ticket, but whether IT can redesign its workflows, operating model, and controls — and serve as the safe backbone for the whole enterprise's AI adoption. The evidence from 2024–2026 is that a minority of firms are achieving genuine redesign while most remain in a productivity-tool phase.

26%
of FS firms run AI in full production today; 65% expect to within 12 months (KPMG, 2026)
27–36%
IT had the largest jump in AI use of any function in six months (McKinsey)
~100k
U.S. advisor shortfall by 2034 — the capacity driver behind WealthTech AI (McKinsey)
230
control objectives in the new Treasury/CRI Financial Services AI RMF (Feb 2026)
14 reference tables · expandable below
Every analytical table from the underlying research report is available below as a click-to-expand data view. The AI use-case map opens by default; the rest are closed to keep the briefing scannable. Use Expand all to open everything at once.
01The Thesis

Three positions this report defends.

IT functions treating AI as a productivity overlay capture some cost benefit but no durable advantage. The functions pulling ahead redesign the work — and in financial services the regulatory floor is rising fast, with a sector-specific control framework now in force.

i.

IT plays two roles at once.

It is both an adopter of AI inside its own work and the backbone that lets every other function adopt AI safely — supplying the data platforms, identity, model governance, and security controls. A firm whose IT house is not in order cannot govern AI elsewhere.

ii.

Reshape, don't replace.

The strongest results — JPMorgan's LLM Suite, Morgan Stanley's AI@MS Assistant, Bank of America's Erica, Citi's developer rollout, Goldman's coding platform — frame AI as augmentation. Klarna's customer-service reversal is the cautionary counter-example.

iii.

The regulatory floor moved in 2025–26.

The Treasury/CRI Financial Services AI RMF (230 controls), the NYDFS third-party-provider letter, amended Reg S-P, and SR 11-7 applied to LLMs have redrawn the lines. Most firms have not yet updated their controls to match.

02Adoption & Context

A vanguard, a middle, and a long tail.

Adoption is stratified. A small vanguard of global banks and asset managers has moved beyond pilots to enterprise scale; an expanding middle tier is hardening governance and data; a long tail consumes AI through vendors. The binding constraint everywhere is data quality, not models.

Where the production maturity is
Report §2.5 · highest-maturity IT use cases
Software engineering & developer productivityGoldman ~20% productivity, 15% fewer bugs; BofA, Citi comparable
Scaled
Service desk & IT operationsTier-0 self-service, triage, knowledge generation
Scaled
Cybersecurity SOC & fraudAI-assisted triage, anomaly detection
Scaled
Document & knowledge workcopilots, meeting summaries, contract review
Scaling
Agentic / autonomous operationscontrolled pilots; new governance required
Emerging

Global institutions with dedicated AI teams and enterprise data infrastructure — JPMorgan, Goldman, Morgan Stanley, BlackRock, Bank of America, Citi — operate in a different category from mid-market firms, which navigate the same transformation with limited governance bandwidth and greater dependence on WealthTech and SaaS vendors. For PE-backed consolidators the integration challenge compounds the adoption challenge: each acquired firm brings different data standards, and AI applied to inconsistent data introduces model risk.

03Operating Model

Centrally-led, federated delivery.

The resolution most large institutions are converging on is a centrally-led hub-and-spoke: a central platform and governance core paired with federated delivery inside the lines of business. The center owns the AI platform, foundation-model contracts, the evaluation harness, retrieval and prompt patterns, the model inventory, and the standards; the spokes own use-case selection and business integration.

01

Fit the model to scale.

For a global bank, a centrally-led platform with federated business AI and an "AI control tower." For a super-regional, a centralized center of excellence. For an RIA or PE-backed consolidator, a lean central AI lead standardizing on a small portfolio of vendor AI, backed by an enterprise acceptable-use policy and an AI inventory.

02

Shadow AI is an operating-model choice.

Prevention is not merely a policy: the approved-tool pathway has to be genuinely easier to use than the unapproved one, or employees route around it.

03

Protect the skills that matter.

Let AI absorb basic troubleshooting, password resets, and routine documentation — but actively protect root-cause analysis, compliance interpretation, and knowledge curation from atrophy. 32% of CTOs already report overdependence on AI for decision-making.

04Data & AI Function

Data is the binding constraint.

The structural change is the move from centralized reporting teams to federated data-product squads on a shared platform. Each squad owns a domain — client, portfolio, advisor, custodian, CRM, planning, market, HR, finance, vendor, or operational data — accountable for data contracts, service levels, quality metrics, and lineage. AI multiplies data quality rather than substituting for it: poor data produces confident, well-documented, poorly grounded output at scale.

What governance frameworks cannot fix
3
Failure modes survive even a fully populated control matrix: ungoverned data; the human-in-the-loop illusion (a reviewer with neither time nor information to overrule the model); and measuring whether controls exist rather than whether AI delivers value. Frameworks are necessary, not sufficient.
New attack surface
RAG
Vector and embedding pipelines create new attack surfaces that traditional controls miss (OWASP LLM08) — a reason the AI function and security must co-own retrieval governance, entitlement-aware retrieval, and embedding-store integrity.
05Infrastructure & Operations

Bounded autonomy, audited everywhere.

AI's operational value is most concrete in I&O — anomaly detection, capacity forecasting, root-cause analysis, remediation drafting, cloud-cost optimization — and its autonomy risk most acute. The governing principle for autonomous action is blast radius: reversible, low-impact actions are reasonable to automate with logging; anything touching production data, security controls, or customer-facing services requires human approval through documented change management.

Service desk
Scaled

Tier-0 self-service · ticket triage · knowledge-article generation. ServiceNow's Now Assist requires a documented reason when a human overrides an AI recommendation — an audit trail supporting regulatory review.

~50%
service-desk call reduction at Bank of America via Erica for Employees
>90%
of BofA teammates use Erica for Employees
Vendor & third-party risk
Hardening

The NYDFS October 2025 third-party-service-provider letter names cloud, AI, and FinTech vendors explicitly: lifecycle due diligence, contractual protections, monitoring, and non-delegation of Part 500 compliance. Consuming vendor-embedded AI concentrates risk rather than transferring it.

~300
banks briefly offline in the C-Edge vendor-incident cascade — the fourth-party risk example
06Cybersecurity

AI versus AI.

AI strengthens SOC triage, anomaly detection, and phishing analysis — while attackers weaponize the same tools for deepfakes and business-email compromise. The OWASP LLM Top 10 (prompt injection, excessive agency, training-data and embedding poisoning) defines the new control surface, and NYDFS warns that some multifactor methods are vulnerable to deepfakes, recommending liveness detection and out-of-band verification for payments and approvals.

Human-in-the-loop on suppression.

Automation bias is the risk in SOC triage: a model that suppresses a novel attack is more dangerous than one that floods the queue. Sample and review suppressed alerts.

Out-of-band for money movement.

Deepfake voice and video impersonation of executives is now a live fraud vector; payment and approval flows need verification that does not depend on the channel being spoofed.

Govern non-human identity.

Agents acquire credentials and act. Least privilege, action logging, and approval gates for agentic actions (LLM06 excessive agency) move from nice-to-have to baseline.

07WealthTech & Capacity

From productivity to capacity.

The most important reframing for a wealth-management CTO is to stop optimizing advisor productivity and start increasing advisor capacity — the number of clients an advisor can serve well — because the binding constraint is demographic, not technological. AI's value in wealth is not a cost-optimization story but a capacity story: it decouples revenue growth from advisor headcount.

Advisor shortfall by 2034
~100k
McKinsey estimates a 90,000–110,000 advisor shortage (30–37% of current headcount) at today's productivity. About 110,000 advisors — 38% of the total, 42% of industry assets — are expected to retire within the decade, against a workforce that grew only ~0.3%/year over the prior ten.
Hybrid engagement by 2029
90%
Gartner predicts 90% of wealth-management firms will have deployed some form of hybrid human-and-digital engagement platform, up from under a third in 2025 (Gartner prediction). The same capacity math applies: hybrid models let a stable advisor base serve more clients.
Morgan Stanley
Global wealth

AI@MS Assistant · Debrief (meeting summaries with consent, written back to CRM) · AskResearchGPT.

>98%
of advisor teams use the AI@MS Assistant
20–80%
document-retrieval coverage shift
JPMorgan
Megabank

LLM Suite · Connect Coach · "Ask David" multiagent investment research (supervisor agent orchestrating structured-data, RAG, and analytics sub-agents, with human-in-the-loop).

200k+
employees reached by LLM Suite; ~4 hrs/week saved (150k+ daily users)
$1–2B
annual AI value cited by Daniel Pinto (mid-2024 estimate later raised toward $2B)
UBS
Global wealth

UBS Red — two Azure-based domain-specific advisor assistants. Illustrates the shift toward domain-specific language models (DSLMs); Gartner predicts ~50% of WM GenAI models will be domain-specific by end-2027 (Gartner prediction).

~30k
client advisors reached
60k+
investment/product documents in the queryable base

The architectural destination is the all-in-one advisor desktop — CRM, onboarding, 360-degree view, planning, portfolio management, order generation, and client communication unified behind one experience. Gartner counts 50+ vendors (Table 14) and predicts 70%+ adoption by 2029 (Gartner prediction). For a CTO the platform choice increasingly determines which AI features advisors get and on whose release cadence — so the vendor's AI roadmap becomes a governance input, and the firm's leverage lies in the integration layer and the data it controls.

All-in-one advisor desktop platform capabilities
Multichannel Tablet Mobile Chat bot Hybrid adviser Advisor desktop platform Risk and compliance CRM Client onboarding 360 degree view Financial planning Investment strategies Model rebalancing Order generation Dashboard Client communication reporting FIX/API Internal and external counterparties ETL ESB/APIs Back office 1 Back office 2 Back office 3
Figure: All-in-one advisor desktop platform capabilities — multichannel front end over an integrated advisor workflow, governed by risk and compliance, connected via FIX/API to counterparties and via an ETL/ESB bus to back-office systems. Adapted from Gartner.
08AI Red Lines

Four tiers of AI permission.

A defensible financial-services IT AI policy distinguishes prohibited, senior-approval, enhanced-review, and normal-review use. The full mapping — with the controlling framework for each — is in Table 11 below.

01Prohibited — no exception

Never permitted

  • Concealing material AI use from auditors, regulators, risk, or compliance
  • Training external models on confidential or source-code data without an executed prohibition
  • Agents changing firewall, identity, network, or permission config without approved change management
  • Deepfake / synthetic-identity tools in access, approvals, payments, or client instructions
  • AI making final credit, employment, investment, or client-treatment decisions without human sign-off
03Enhanced review

Technology / security review

  • RAG over sensitive client, portfolio, cyber, HR, or source-code data
  • Agentic workflows touching production data or external APIs
  • Customer data in evaluations or fine-tuning
  • Unapproved public AI tools (default prohibited; approved enterprise tooling required)
04Normal review

Allowed with standard QA

  • Productivity copilots within approved data domains
  • Code completion in approved IDEs with secure review enforced
  • Knowledge retrieval over approved internal bases
  • Meeting summarization with client consent and CRM logging (the Morgan Stanley pattern)
09Regulatory Floor

The floor rose in 2025–26.

No single "AI law" governs financial-services IT; instead a stack of existing and new authorities applies. The full mapping to IT implications is in Table 10. The headline shift is a sector-specific control framework and a sharpened third-party-risk posture.

Feb 2026Treasury/CRI
U.S. Treasury & Cyber Risk Institute
Financial Services AI Risk Management Framework
NIST-aligned, 230 control objectives mapped to AI-adoption stage, developed with 100+ institutions. Voluntary but examination-relevant — the most significant FS-specific AI-governance development in the period.
Oct 2025NYDFS
NY Dept. of Financial Services
Third-party service-provider letter (23 NYCRR 500)
Lifecycle third-party risk explicitly naming cloud, AI, and FinTech providers; senior-officer accountability; Part 500 compliance cannot be delegated to a vendor.
2024–26SEC/FINRA
SEC & FINRA
Amended Reg S-P · FINRA 2025/26 reports · RN 24-09
Service-provider oversight, incident response, and 30-day breach notification; GenAI/agentic AI supervised under existing rules with sharpened third-party focus.
2011/nowFed/OCC
Federal Reserve / OCC
SR 11-7 model-risk discipline applied to AI/LLMs
Validation, inventory, and ongoing monitoring expectations extend to ML, AI, and large language models used in credit, AML, and investment contexts.
Aug 2024EU
European Union
EU AI Act — high-risk obligations
Credit scoring, insurance pricing, and employment AI are high-risk. Annex III dates were deferred under the May 2026 Digital Omnibus political agreement — pending final adoption, so re-verify before circulation.
10Maturity Model

Five levels. Most firms sit at 2–3.

The diagnostic companion to the use-case map. The leaders profiled in the case studies operate at Level 4; only the largest approach Level 5 in selected domains.

Level 01
01
Ad-hoc experimentation
Employees use public GenAI tools with no inventory or policy. Typical of small RIAs and early-stage fintechs.
Level 02–03 · Most firms
2–3
Functional pilots — scaled use cases
An AUP and AI inventory begun; then an internal LLM platform, SDLC and service-desk copilots, event intelligence, a vendor-AI inventory, and model-risk integration. Mid-size to super-regional.
Level 04 · Leaders
04
Workflow & operating-model redesign
Product/platform operating model, data products, advisor copilots, agentic workflows with controls, an AI control tower, explicit red lines, eval-driven release. Top-20 banks, global asset managers, large wealth platforms.
Level 05 · Largest only
05
AI-native, resilient, product-platform IT
Agents augment most workflows; non-human identity governance, continuous control monitoring, resilience-by-design, full SR 11-7 / NIST / EU AI Act alignment. Selected domains only.
Self-diagnostic: if you cannot point to a workflow structurally rebuilt around AI — not just augmented — and to new roles created in the past 18 months, you are at Level 2–3 regardless of tool count. Progress has come from adding capability, not subtracting people.
11CTO Action Agenda

A 12-to-24-month plan.

Four phases from foundation to operating-model and resilience — sequenced so governance and data precede agentic scale.

Months 0–3
Foundation
01
Stand up the AI Governance Office and AI Control Tower. Confirm three-lines roles.
02
Inventory all AI in use — in-house and vendor-embedded — and merge with the model inventory; issue an AUP aligned with the FS-ISAC framework.
03
Block unapproved external GenAI at egress and provide an approved alternative that is genuinely easier to use.
04
Map the regulatory perimeter (FFIEC AIO, SR 11-7, Reg S-P, FINRA, NYDFS, FS AI RMF, EU AI Act) and assign owners.
Months 3–9
Platform & risk hardening
05
Launch or consolidate an internal LLM platform with an evaluation harness, prompt and retrieval patterns, and audit logging.
06
Productize the top three to five IT use cases (SDLC copilot, service-desk copilot, event intelligence, vendor-management AI, knowledge management).
07
Implement OWASP LLM Top 10 controls and LLM red teaming; begin shifting two or three priority data domains to data products.
08
Renegotiate the top-20 vendor contracts on data use, training, audit rights, exit, and sub-processors; enforce the red lines technically.
Months 9–18
Workflow redesign
09
Move priority workflows from pilot to scaled production: advisor desktop, software engineering, service desk, SOC, vendor management, PMO.
10
Establish a recurring vendor-AI evaluation cadence across Workday, ServiceNow, SAP, Salesforce, Aladdin, Orion, Addepar, Envestnet, and surveillance tooling.
11
Pilot agentic workflows with strict bounded autonomy, rollback, and human approval thresholds; keep agents out of production change management without governance.
12
Mature model-risk management to cover generative AI; extend SR 11-7 validation.
Months 18–24
Operating model & resilience
13
Transition IT to a product and platform operating model with embedded controls and continuous control monitoring.
14
Address EU AI Act high-risk obligations (credit scoring, insurance pricing, employment AI) ahead of the deferred date, re-verifying the timeline.
15
Run a skill-shift program — AI fluency for engineers, data-product management, AI-risk roles — and deliberately protect the junior-talent pipeline.
16
Hold an annual board AI-strategy review using control-tower metrics.
12 · So What

What CTOs should do differently now.

Stop benchmarking AI adoption by tool count or pilot count. Start benchmarking by the question the evidence converges on: how much of the IT function's work has been redesigned around AI, and can the firm show — to an examiner — who or what made each consequential change and on what authority?

By 2027, the difference between IT functions that thrived through the AI transition and those that did not will not be model access — every firm will have the same frontier models, the same clouds, and many of the same vendors.

The difference will be (1) how decisively the operating model moved to product-and-platform with centrally-led governance and federated delivery; (2) whether data quality, model risk, vendor risk, and the red lines were built in upstream rather than bolted on; and (3) whether IT became — or refused to become — the safe backbone for the enterprise's AI adoption, not merely an adopter of its own tools.

In wealth specifically, the winning CTOs treat AI not as a cost-optimization story for the CFO but as a capacity story: the way a stable or shrinking advisor base serves more clients well, against a ~100,000-advisor shortfall that technology alone cannot close.

End of briefing · The CTO's AI Decision · IT in Financial Services
Your future is our focus.™
14 Reference Tables · Expandable

The full evidence, on demand.

Every analytical table from the underlying research is below as a click-to-expand data view. The AI use-case map is open by default — it is the single most useful artifact for prioritization. The other thirteen are closed to keep the briefing scannable; open one, several, or all at once.