AI is transforming investment banking in 2026, but it is restructuring the profession rather than replacing bankers outright. Major banks — JPMorgan, Goldman Sachs, Morgan Stanley and Citi — now run AI at enterprise scale, automating modelling, due diligence and reporting. Junior analyst roles are shrinking fastest, while client-facing advisory and deal judgement remain firmly human-led. For UK firms navigating the same shift, Whitehat's AI consultancy and implementation service connects AI adoption to measurable commercial outcomes.
$177M
Avg AI spend per bank, next 12 months (Q1 2026)
~60%
JPMorgan workforce using AI in production
~25%
Banking work hours automatable (Goldman)
54%
Financial jobs with high automation potential (Citi)
Sources: Banking Dive CIO survey Q1 2026; JPMorgan technology strategy 2026; Goldman Sachs research; Citigroup workforce automation report.
How Is AI Being Used in Investment Banking in 2026?
In 2026, the largest investment banks have moved well past pilot projects and now run AI across core functions — deal origination, due diligence, financial modelling, trading, compliance and client service. This is enterprise-scale deployment, not experimentation. Sector-wide, projected AI spend reached an average of $177 million per bank for the coming 12 months in a Q1 2026 CIO survey — up 33% quarter-on-quarter — and hyperscale AI capital expenditure across financial services is estimated at roughly $700 billion in 2026, around five times the level of five years earlier.
Each of the tier-one banks has built its own AI platform and pushed adoption across the workforce:
| Bank | AI Platform / Investment | Scale & Impact (2025–2026) |
|---|---|---|
| JPMorgan | LLM Suite; ~$20bn 2026 tech budget | 500+ AI use cases; ~60% of workforce using AI in production; ~4 hours/week saved per user (≈600,000 hours firm-wide) |
| Goldman Sachs | ~$6bn 2026 tech; Anthropic (Claude) partnership | Devin AI agent across 12,000 developers; 3–4× engineering productivity; AI in compliance, equity research, credit risk |
| Morgan Stanley | OpenAI partnership; AI @ Morgan Stanley Assistant | ~280,000 hours saved via code translation; ~98% of advisers use it daily; 30–40% less admin time |
| Citigroup | Firm-wide generative-AI tooling | Rolled out to 182,000 employees in 84 countries; >70% adoption; ~100,000 developer hours saved per week |
Sources: Observer (Goldman/Anthropic, Feb 2026); Bloomberg; Morgan Stanley Insights 2026; Banking Dive 2026; Fortune 2026.
Underneath the platforms, AI is concentrated in a handful of high-volume workflows. In compliance, HSBC's anti-money-laundering AI screens around a billion transactions a month and flags suspicious activity far more accurately than legacy systems. In everyday productivity, Barclays has put roughly 100,000 employees on Microsoft 365 Copilot — one of the largest workplace AI deployments in financial services. And in deal execution, McKinsey documented a leading bank cutting investment-brief production from nine hours to about 30 minutes — a more-than-90% reduction that resets expectations for analyst turnaround.
The pattern is consistent: AI is being embedded as core infrastructure spanning strategy, delivery and reporting, not bolted on as a departmental tool. The same restructuring is visible across professional services more broadly — our analysis of AI's impact on the consulting industry shows the Big Four following an almost identical playbook of platform investment, mandated adoption and outcome-based pricing.
Will AI Replace Investment Bankers?
No — AI will not replace investment bankers wholesale, but it is permanently restructuring the profession. The strongest evidence comes from the banks' own leaders. Goldman Sachs CEO David Solomon says AI will "reshape banking work, not wipe it out," calling fears of a job apocalypse overblown while estimating that around 25% of work hours could be automated — time he expects to shift toward higher-value client work. JPMorgan's Jamie Dimon is blunter: "I think it will reduce our jobs down the road," but adds the bank will hire more AI specialists and fewer traditional bankers, with remaining bankers made more productive.
The nuance matters. Fortune's 2026 investigation concluded that the AI finance "takeover" is "largely smoke and mirrors" — banks are overwhelmingly using AI to augment roles rather than eliminate them. The clearest illustration is the analyst multiplier: a single first-year analyst supervising AI can now produce work that once required three analysts. That is not replacement; it is a radical productivity increase that changes how many junior seats a bank needs, not whether it needs bankers at all.
The consensus among analysts is that AI is replacing tasks, not entire front-office roles. Routine modelling, data cleaning, first-draft pitch books and document review are being automated; client relationship management, negotiation, risk judgement, deal structuring and bespoke strategic advice remain human-led. A 2026 sector analysis put it plainly: "AI is not replacing investment bankers at scale, but it is restructuring roles." Morgan Stanley's research goes further, arguing AI will be a net positive for banks — with current job losses concentrated among junior and entry-level positions rather than the advisory franchise.
Key Takeaway
The "will AI replace investment bankers?" question is the wrong frame. AI is compressing the traditional career pyramid — automating the transactional lower rungs while concentrating value in judgement, relationships and deal-making. The bankers most at risk are those whose work is purely execution; the most protected combine domain expertise with the ability to direct and validate AI.
Which Investment Banking Jobs Are Most Affected by AI?
Entry-level analytical work is by far the most exposed. Goldman Sachs research suggests AI could automate about a quarter of current banking work hours, with junior tasks — financial modelling, note-taking, spreadsheet analysis and deck formatting — "most exposed." A Citigroup report found 54% of financial jobs have high automation potential. McKinsey's QuantumBlack leadership told Fortune that banks are cutting junior analyst classes by up to two-thirds, while sourcing roughly 62% of their AI talent from those same cohorts — redeploying juniors into technical and AI roles rather than traditional modelling.
Headcount actions across the majors show the same signal — productivity rising while entry-level intake falls:
| Function | Declining Demand | Rising Demand |
|---|---|---|
| Financial modelling | Manual Excel model building | AI model validation, assumption oversight |
| Due diligence | Manual document review & extraction | Risk interpretation, anomaly judgement |
| Pitch books | Deck formatting, comps tables | Narrative, positioning, client tailoring |
| Client advisory | Low — limited automation | High — relationships, negotiation, judgement |
| Technology fluency | N/A (new requirement) | Prompt engineering, AI workflow design |
Sources: Citigroup workforce automation report; McKinsey/QuantumBlack via Fortune 2026; Morgan Stanley thematic research 2026.
The headcount evidence is stark. Morgan Stanley cut around 2,500 jobs after posting record 2025 revenue of $70.6 billion; Citigroup has committed to reducing its global workforce by roughly 20,000 roles; JPMorgan's Jamie Dimon says AI has let the bank cut jobs by as much as 40% in some units; and Goldman has linked 1,000+ role reductions to AI productivity. Across the largest US banks, projections point to up to 200,000 banking jobs lost over three-to-five years as AI absorbs back- and middle-office work. Yet an EY survey of 240 financial-services CEOs found 60% expect AI investment to maintain or increase headcount in 2026 — evidence that many firms see AI as a way to avoid further staff-cost growth rather than to slash roles immediately.
Is Investment Banking Still a Safe Career in the Age of AI?
Investment banking remains a viable career, but the safe ground has moved. The roles most insulated from AI are those built on what AI cannot replicate at scale: client relationships, commercial judgement, negotiation, sector expertise and the ability to structure and close complex deals. The roles most at risk are those defined by repeatable, rule-based execution. In practice, that means the career ladder is being squeezed in the middle — fewer purely transactional junior seats, more demand for people who pair domain knowledge with genuine AI fluency.
More protected
Senior advisory and coverage; M&A deal structuring; capital raising; client relationship management; risk and credit judgement; roles combining sector expertise with AI oversight.
More exposed
Purely transactional junior analyst work: manual modelling, deck formatting, data processing, first-pass document review, routine reporting and reconciliation.
Technical fluency with AI tools is fast becoming table stakes. You do not need to be a machine-learning engineer, but you do need to prompt effectively, validate AI outputs and spot where models fail. The bankers best positioned right now are often those with five-to-ten years of experience who have added Python or prompt-engineering skills on top of deal expertise. Demand is also rising for entirely new roles — AI strategists, prompt engineers and AI-human workflow designers — that sit at the intersection of technical and domain knowledge and frequently command a premium over equivalent analyst positions.
Client demand reinforces why relationship-led roles are the most defensible. Even as AI advances, only around 6% of UK clients seeking investment advice would rely on an AI platform alone, while roughly 34% are comfortable with human advisers using AI tools. Among affluent households, about 80% say they would pay a premium for human advice over an exclusively digital service. The human relationship is not a nostalgic preference — it is what clients still pay for.
What Productivity Gains Are Banks Achieving with AI?
The productivity gains driving all this are real and measurable. JPMorgan's CFO has said AI effectively doubled the productivity impact of technology, with employees using its internal LLM Suite saving roughly four hours a week — about 600,000 hours across the firm. Morgan Stanley estimates AI will deliver 20–50% productivity gains across wholesale banking, trading and operations over five-to-ten years, translating into an ≈18% uplift in pre-tax income once embedded. Goldman reports 3–4× engineering productivity from AI coding agents, and Citi has saved around 100,000 developer hours a week through automated code review.
A Morgan Stanley survey of AI-adopting companies captured the dual trend precisely: an 11.5% average net productivity increase alongside a 4% net headcount decline over twelve months — productivity-linked job thinning rather than mass layoffs. McKinsey and Accenture reviews suggest AI could trim banking costs by up to 20% while improving software engineering, risk management and customer service. The strategic implication for firms procuring advisory or technology partners is the same one reshaping consulting: efficiency gains of this scale make traditional time-billed pricing increasingly hard to defend, and sophisticated buyers are beginning to demand outcome-linked models.
The prize is large enough to explain the spending. McKinsey estimates AI could unlock $200–340 billion in annual value for the financial-services sector, and EY research shows 47% of banks have now fully deployed generative AI — up from just 10% in 2023. In advisory specifically, Goldman Sachs has projected a 33–41% uplift in M&A fees from AI-driven productivity, as faster diligence and modelling let teams run more deals in parallel without proportionally more headcount.
How Should Bankers and Firms Respond to AI?
Whether you are an individual banker or a financial-services firm, the response to AI operates on two levels — personal skills and organisational capability. The practical steps below are drawn from what the leading adopters are already doing.
1
Build genuine AI fluency, not awareness
Learn to prompt, validate and supervise AI outputs on real banking tasks. Fluency — knowing where models help and where they fail — is now a baseline expectation, not a differentiator.
2
Move up the value chain
Concentrate on the work AI cannot do — client relationships, deal judgement, negotiation and sector depth. Let AI absorb the modelling and formatting; own the interpretation and the relationship.
3
Invest in governance before scaling AI
In a regulated sector, unmanaged AI is a liability. Establish data protocols, model validation, audit trails and review processes before widening deployment — not after an incident.
4
Reskill juniors, don't just cut them
The banks sourcing ~62% of AI talent from junior cohorts show the smarter path: redeploy entry-level staff into AI-augmented and technical roles rather than simply shrinking the intake.
5
Tie AI to measurable outcomes
Define the business result — hours saved, deals-per-banker, cycle time, cost-to-income — before deploying tools, so investment is judged on impact rather than activity.
Navigating AI adoption in a regulated financial-services firm?
We help banks, fintechs and advisory firms build AI strategies and inbound marketing engines connected to measurable outcomes — not tooling checklists.
Explore AI Consultancy & ImplementationWhat Does the UK Regulatory Landscape Mean for AI in Banking?
UK regulators are actively shaping how banks deploy AI, and the environment is supportive rather than light-touch. The FCA runs three parallel programmes — a Supercharged Sandbox providing computational resources for AI testing, AI Live Testing for controlled real-world trials, and an AI Sprint convening industry and regulators on specific governance questions — alongside the Mills Review of AI's impact on retail financial services. Around 75% of UK financial firms already use AI in some capacity, according to the Bank of England and FCA.
⚠️ Regulatory Risk
Banks with EU operations face a dual compliance burden: the EU AI Act's high-risk obligations — covering AI-based creditworthiness assessment and risk pricing — begin full enforcement in August 2026. UK firms serving EU clients must satisfy both regimes.
As AI-generated content and analysis become routine, expect stricter oversight of how banks use AI in client communications and advice. Firms deploying AI without robust governance — model validation, audit trails, human oversight — face enforcement and reputational risk that far outweighs any efficiency saved.
This regulatory reality creates a moat for compliance-aware operators. Generic AI deployment causes delays in a sector where every claim requires substantiation and every model needs governance. Firms that understand both the AI landscape and the specific rules of their sub-sector — institutional versus retail, wealth versus insurance — will move faster and more safely. It is the same governance-first pattern we see across regulated industries, and a core part of how we approach AI consultancy and implementation for UK B2B clients.
Frequently Asked Questions — AI in Investment Banking
Will AI replace investment bankers?
No — AI will not replace investment bankers wholesale, but it is substantially restructuring the profession. Bank leaders including Goldman's David Solomon and JPMorgan's Jamie Dimon expect AI to automate around 25% of work hours and reduce some roles, while making remaining bankers more productive. AI replaces tasks — modelling, document review, first-draft pitch books — not the client relationships, negotiation and deal judgement that define senior banking.
Which investment banking jobs are most at risk from AI?
Junior analyst and associate roles focused on financial modelling, deck formatting and data processing are most at risk — the most transactional, rule-based work. Citigroup found 54% of financial jobs have high automation potential, and banks are reportedly cutting junior analyst classes by up to two-thirds. Senior, relationship-led and judgement-heavy roles are far more insulated.
Is investment banking still a safe career in the age of AI?
Yes, but the safe ground has shifted. Careers built on client relationships, deal structuring, negotiation and sector expertise remain strong, while purely execution-based junior work is shrinking. The best-positioned bankers pair deal experience with genuine AI fluency. An EY survey found 60% of financial-services CEOs expect AI to maintain or increase headcount in 2026, so the outlook is restructuring rather than collapse.
How much are investment banks spending on AI in 2026?
Spending is at record levels. JPMorgan's 2026 technology budget is around $20 billion and Goldman Sachs has earmarked roughly $6 billion, with large portions going to AI. A Q1 2026 CIO survey put average projected AI spend at $177 million per bank for the coming year, up 33% quarter-on-quarter, while hyperscale AI capex across financial services is estimated near $700 billion — about five times the level of five years earlier.
Can AI take over M&A and deal-making?
AI is accelerating the mechanics of deals — due diligence, financial modelling, market research and first-draft materials — with some banks reporting brief-production time cut by over 90%. But deal-making itself depends on relationships, negotiation, judgement on risk and bespoke structuring, which remain human-led. AI handles higher deal volumes and faster cycle times; it does not replace the banker who wins and closes the mandate.
How should investment bankers prepare for AI?
Build real AI fluency — learn to prompt, validate and supervise AI on genuine banking tasks — and move up the value chain toward client work, judgement and deal-making that AI cannot replicate. Develop deep sector expertise, cultivate client relationships, and treat AI as a tool to multiply your impact. Bankers who combine five-to-ten years of experience with prompt-engineering or data skills are among the best-placed in the market.
Leading a Financial-Services Firm Through AI Change?
We help banks, fintechs and advisory firms build AI strategies and inbound marketing engines that attract the right clients and connect directly to commercial outcomes. We understand both the AI landscape and the financial-services buyer journey.
Clwyd Probert
Clwyd Probert is the founder of Whitehat, a London-based SEO and inbound marketing agency and HubSpot Diamond Solutions Partner since 2016. He advises B2B companies in regulated industries — including investment banks, fintech platforms and advisory firms — on SEO, AI search optimisation and content strategy that delivers measurable pipeline growth.