What gets automated, what gains value, and what it means for careers, banks and LPs
For decades, the first years of an investment banking or private equity career followed a fixed formula: build the comps, update the model, format the deck, draft the first memo. Generative AI now performs much of that work in minutes. The question for professionals, hiring managers and LPs is no longer whether AI changes the junior role, but how.
The headline: tasks are changing faster than titles
The evidence is mixed, and anyone claiming certainty should be treated with caution. Some industry reporting suggests banks are shrinking junior analyst intakes by as much as two-thirds, according to McKinsey’s QuantumBlack, and leaders such as JPMorgan’s Jamie Dimon and Citi’s Jane Fraser have acknowledged that AI will eliminate or make redundant some roles. Yet Fortune’s own investigation concluded that the AI takeover of finance is largely overstated, with banks mostly using AI to augment roles rather than eliminate them. Both can be true: tasks are being automated much faster than roles are being abolished, and that changes how many junior seats a team needs.
Task by task: what AI is taking, sharing and leaving
| Task | AI impact | What remains human |
|---|---|---|
| Comparable company and precedent screens | Largely automated | Choosing the right peer set and explaining outliers |
| Formatting pitch books and decks | Largely automated | Storyline and what to leave out |
| First drafts of memos (CIM summaries, IC memos) | Heavily assisted | Deciding what matters; challenging the narrative |
| Financial model population and scenarios | Heavily assisted | Whether the assumptions make sense, and what results mean |
| Deal sourcing and target screening | Heavily assisted | Judging fit, reputation and timing |
| Document review in diligence | Heavily assisted | Interpreting red flags, negotiating consequences |
| Negotiation, structuring, client trust | Little impact | Almost everything |
The pattern is consistent. As one analysis puts it, AI can populate models, organise information and run scenarios, but analysts still need to decide whether the assumptions make sense and what the results mean.
What this looks like in private equity
PE teams are moving the same way. An Affinity survey of 275 private equity and venture capital professionals found that the share of firms using AI inside the investment decision itself, not just around it, rose from 13% to 28% in a year. Screening teasers, ranking targets against a buy-box and synthesizing expert calls are now routine uses of AI.
The risks are real. Practitioners point to hallucinations, overreliance by staff, data breaches and non-compliance as the main risks firms must manage. An analyst who cannot verify an AI-generated output is a liability, not a productivity gain.
Skills that gain value
- Judgment and verification. Reviewing, testing and correcting machine output replaces building from scratch. This requires strong fundamentals; you cannot audit a model you could not have built.
- Relationships and communication. When production work shrinks, juniors are pulled into client interactions earlier in their careers. arvya
- Structuring. Capital structure, deal terms and the mechanics behind a transaction (see our guides on valuation multiples and due diligence) are where human expertise compounds.
- AI fluency. This is becoming a hiring criterion. UBS is reportedly requiring candidates for junior banking roles to prove AI proficiency before being hired.
The apprenticeship paradox
There is a structural problem the industry has not solved. If AI absorbs entry-level work, where do future associates and VPs learn their craft? McKinsey’s QuantumBlack data suggests banks source 62% of their AI talent from junior analyst cohorts, which means cutting those classes may weaken the firms’ own AI pipelines. Today’s analysts are tomorrow’s MDs. Firms that automate away the learning curve without replacing it risk a leadership gap a decade from now.
What this means for each audience
For students and junior professionals: master the fundamentals first, then add AI fluency. Technical interviews will keep testing whether you understand a model, not whether you can prompt one. Our guide on CFA, CAIA or MBA covers how credentials still signal competence.
For banks and GPs: redesign training around review, judgment and client exposure. Build governance for data confidentiality and model risk before scaling tools.
For LPs: add AI governance to manager due diligence. Ask how a GP uses AI in sourcing and underwriting, who verifies outputs, what data leaves the firm and whether efficiency gains show up in fees or returns. An efficiency edge is only an edge if it is controlled.
Conclusion
AI is not eliminating the banker or the PE investor. It is removing the production layer that used to teach them. The professionals who thrive will combine solid technical foundations, sound judgment and the ability to direct and verify AI.
FAQ
Will AI replace investment bankers? Not at scale in the near term. Evidence points to tasks being automated rather than front-office roles disappearing, but junior intake and team composition are changing.
Which investment banking tasks will AI automate first? Repetitive, rule-based work: comps screening, deck formatting, model population and first-pass document summaries.
What skills should a junior banker or PE analyst develop? Strong accounting and valuation fundamentals, judgment in reviewing outputs, communication and structuring skills, and practical AI fluency.
This article is for informational purposes only and does not constitute investment or career advice.