In 2023, a McKinsey report estimated generative AI could automate 30% of the tasks performed by financial analysts. Two years later, the estimate feels conservative. Tools like BloombergGPT, Aiera, FinChat, and the AI features baked into Excel and Google Sheets now handle model building, scenario analysis, pitch book drafting, and first-pass due diligence summaries. And yet financial analyst job postings have stayed roughly flat, and senior analyst comp at top firms is rising. Understanding why tells you where the role is going.
What AI has already absorbed
A modern finance workflow has large chunks that are now essentially automated:
- Three-statement models. AI can build a credible starter model from a 10-K in under an hour. The formula work, the linking, the basic drivers — all table stakes now.
- Comparable company analysis. Pulling comps, calculating multiples, building football field charts — largely automated via tools like Finbox and AlphaSense.
- Earnings transcript synthesis. Summarizing a quarter's worth of calls and flagging tonal changes used to be a junior analyst's week. Now it's a half-hour.
- Pitch book formatting. Branded decks, chart generation, layout consistency — software does it.
- First-pass due diligence. AI can read a data room and flag obvious issues far faster than a human.
Goldman Sachs, JPMorgan, Morgan Stanley, and most bulge brackets have rolled out internal AI tools (GS's "Socrates," JPM's "LLM Suite") and publicly reported 20–40% time savings on first-year analyst tasks. Analyst headcount at some banks has stayed flat despite deal volume rising — the productivity gain is being absorbed into output, not layoffs, at least so far.
What AI is still bad at
Judgment about assumptions. A DCF is only as good as the revenue growth assumption and the terminal multiple. An AI will happily generate a model with plausible-looking inputs that are completely wrong for your specific company, industry cycle, or competitive context. Knowing that the consensus 15% growth number is soft because channel checks suggest inventory destocking is a judgment an analyst earns through years of following a sector.
Talking to management. On every serious deal, the biggest signal isn't in the public filings — it's in the body language of the CFO in a management meeting, the side comments from a VP of sales, the pause before answering a question about churn. AI does not attend the meeting.
Stakeholder work. In corporate finance and FP&A roles, the real job is partnering with business leaders, pushing back on unrealistic forecasts, and negotiating resource allocation. In M&A and private equity, it's navigating deal dynamics across buyers, sellers, and lenders. None of this is in any training set.
Context-heavy storytelling. A good investment memo or board deck tells a coherent story about why a thesis works. AI produces the words; the analyst produces the view. The difference is what senior partners actually pay for.
Where the comp is going
Levels.fyi and Emolument data from 2025 tell a consistent story:
- Entry-level analyst roles in IB: comp flat in real terms since 2023.
- Senior associate and VP-level roles at top funds: up 8–15% YoY.
- Corporate development and strategic finance at AI-exposed companies: up 20%+.
- PE and hedge fund buy-side roles: continuing the long-term upward trend, with top-bucket analysts at quant-heavy funds seeing outsized gains.
The premium has consolidated around analysts who can combine financial judgment with technical fluency — people who can reason about unit economics of AI products, LLM cost structures, and compute margins as well as they can build a LBO model.
Five things to do this quarter
-
Master AI tools, then audit their output ruthlessly. Use Aiera, AlphaSense, and whatever your firm has. But never trust a model you didn't personally sanity-check. The analysts getting fired are the ones who submit AI output unchecked.
-
Own a sector. Generalist analysts are losing ground fast. Specialists who know the unit economics, competitive dynamics, and management teams of a specific industry are harder to replace and better paid. Pick one and go deep.
-
Learn Python and SQL properly. Excel is not enough anymore. Modern analysts read and write Python for data work, and SQL for pulling from warehouses. The premium for this combination has never been higher.
-
Get into rooms. Sitting in management meetings, on board calls, in deal negotiations — these are the high-value reps that build judgment. If your current role keeps you in Excel all day, it's time to push for more exposure or change roles.
-
Build a view. The analysts who get promoted have opinions — on companies, sectors, deals. Start writing yours down, at least privately. Being able to articulate a thesis clearly is the skill that moves people from associate to VP.
What this means for your applications
Finance resumes in 2026 need to lead with deal sheets and quantified impact, not responsibilities. Specific deals closed, models built that informed decisions, analyses that changed strategy. The financial analyst resume example shows the pattern, and the financial analyst cover letter example anchors it in a narrative that holds up to case interviews.
Interview prep has also gotten harder: expect technical modeling questions alongside AI strategy and business-case work. The financial analyst interview questions guide covers the current shape. For comp context, the financial analyst salary guide has 2026 ranges by level and firm type.
The honest conclusion
Financial analysts are not being replaced wholesale. But the role is being compressed. The mechanical work — modeling, formatting, comps, summarization — is shrinking as a share of the job. The judgment work — assumption-building, sector expertise, stakeholder navigation, thesis construction — is expanding as a share of the paycheck.
Analysts who double down on the judgment layer are going to have good careers. Analysts who defined themselves by their Excel speed are going to find the market has moved on.
