Can AI (ChatGPT, Claude, etc.) Replace Junior Analysts?

I’ve been hearing a lot of talk lately that AI are ready to replace junior analysts, especially in research-heavy roles — so I decided to put that to the test.

I ran 6 of the top AI through a real financial research assignment:

  • GPT-5
  • Claude Sonnet 4.5
  • Gemini 2.5 Pro
  • Manus
  • Pokee AI
  • Skywork

The task:

Create a professional training guide for the U.S. EXIM Bank Single-Buyer Insurance Program (2021–2023) — a document that’s actually usable for training financial advisors and screening clients.

Results summary:

  • Speed: Gemini 2.5 Pro was fastest (~7 min); others took 10–15 min.
  • Quality: Claude Sonnet 4.5 and Skywork produced professional-grade results; GPT-5 underperformed.
  • Instruction following: Claude understood the assignment best.
  • Sources: Skywork had the most relevant, verifiable citations.

TL;DR:
Claude Sonnet 4.5 and Skywork delivered outputs close to what a human analyst might produce. The rest — including GPT-5 — still struggled with coherence, compliance, or practical structure.

At least for this open-ended, real-world research tasks, AI shows obvious limitation..

Attached below (images 2–7) are anonymized outputs from each agent.
Which one looks most professional to you?
Would you trust any of these as a junior analyst’s report?

Drop your thoughts below 👇

22 Comments
 
Most Helpful

I think you are missing the general purpose and reason why an IB analyst is necessary. The reason why IMO Ai won't takeover analysts is that a huge part of the job is to develop the skill set to move up to associate/VP. While a hefty amount leave after 2 years, banks rely on those who've grinded it out in the junior levels to lead the next generation. Theoretically, even if AI could perfectly replicate a human at junior levels, that would be great for... 5 years but after the 5 year mark, where would you find Sr associates or Vps who actually know how to do shit. Unless you think the entire work chain will be replaced, analysts should in general be fine. 

You have to realize, despite recruiting for these jobs being super tough, the actual day to day is a lot less tough (in terms of the mental strain of the work, obviously the hours are what is the killer). Aligning logos and spreading comps doesn't take a genius, as it shouldn't but it provides a learning ground for juniors to help out while also picking up the higher level skills they will need to progress while also understanding the life of what it feels at the bottom. 

 

100% agree! This test was only designed to evaluate one specific aspect of an analyst’s role — the ability to conduct research and produce written reports. There are many other dimensions to being an analyst, especially as they progress toward associate or VP roles, that AI simply can’t replicate.

 

Did you review the work to confirm it wasn’t all hallucinated nonsense? Being hyperbolic, obviously, but that’s a frequent issue I run into with more esoteric content.

 

As a freshman at a target school who aspires to work on Wall Street, this already scares me. I know with time AI is only going to get more powerful. I feel it is a matter of when, not if, AI reduces the amount of entry level high finance jobs. I hope this is not the case but it seems like a sad reality. Maybe I switch to Pre-med lol. idk. 

 

Can it? Yes?  Will it be the general models you tested here? No. 

People keep being obsessed with this doomer view of AI taking all of the jobs.  In reality that will not happen.  AI is a tool that allows humans to optimize their time to focus on higher value human optimized work. 

Also, anyone who thinks that more jobs will not be the result of this are the same ones who were complaining about tractors taking over manual farm work.  All of those people clearly didn't end up in some ditch somewhere after dying of stravation.

 

It will most heavily impact all of the 'marginal' jobs and/or seats across the industry. You won't need the 10th-12th analysts who probably were going to be let go anyway, as AI will be able to take some of the work you'd send their way. Same with other areas of the financial services industry - it will cut at the margins, and give leverage to those who are able to utilize it effectively.

There's a huge amount of applications for AI that we've seen in our firm, and we are just getting started. The risk to most of these timelines are the legal, privacy, compliance, risk, and regulatory considerations of these. Especially for big banks or large institutional managers - they will be slower to adopt, as the 'risk' of saying "Whoops - my AI model didn't quite calculate my risk limits, and I mis hedged my balance sheet' or 'oh yeah, my AI reporting bot screwed up your performance calcs' - is a pretty big issue. Anecdotally, at the institutional level, everyone wants to see or hear that you are looking at it, testing it, starting to use it - but ultimately, they want the human sign off and/or touch to get it across the finish line.

Where you will see it most is smaller, privately held, more nimble firms who has less beau acracy to speed the development. Those are where you'll see the most interesting applications day to day, and then will filter out from there.  

 

Yeah, I see many new AI companies focusing on compliance and legal right now. I think we’re still very early in the process, and we’ll have to see what happens next.

 

Culpa repellendus consequatur quae unde at numquam cumque dolorum. Quisquam et aut temporibus enim sequi doloremque rem ipsa.

Ut eum minima et vel fugiat molestias. Unde laudantium exercitationem iste aut accusamus. Sit aut sequi sed labore dolorem enim. Quae accusantium quas consequatur voluptas omnis similique. Officia aut tenetur aut esse maiores ullam modi.

Impedit corporis accusantium numquam suscipit itaque nam fugiat. Quibusdam eos quam ea reprehenderit sit et et. Quia voluptatum inventore quia necessitatibus dolor optio magnam. Magni non et iure. Aliquid ut numquam dolore esse. In placeat rem ut. Fugit labore possimus sed ut eos. Omnis placeat dicta laudantium voluptas neque et suscipit.

Career Advancement Opportunities

June 2026 Investment Banking

  • Evercore 01 99.4%
  • Moelis & Company 01 98.8%
  • JPMorgan 01 98.2%
  • Guggenheim Partners 01 97.7%
  • Morgan Stanley 07 97.1%

Overall Employee Satisfaction

June 2026 Investment Banking

  • Moelis & Company No 99.4%
  • Morgan Stanley 02 98.8%
  • Evercore 01 98.2%
  • BMO Capital Markets 12 97.6%
  • Banco Santander 01 97.1%

Professional Growth Opportunities

June 2026 Investment Banking

  • Evercore 01 99.4%
  • Moelis & Company 01 98.8%
  • Morgan Stanley 05 98.2%
  • JPMorgan No 97.7%
  • BMO Capital Markets 12 97.1%

Total Avg Compensation

June 2026 Investment Banking

  • Vice President (14) $434
  • Associates (43) $259
  • 3rd+ Year Analyst (8) $210
  • 2nd Year Analyst (22) $179
  • Intern/Summer Associate (13) $156
  • 1st Year Analyst (77) $151
  • Intern/Summer Analyst (71) $101
notes
16 IB Interviews Notes

“... there’s no excuse to not take advantage of the resources out there available to you. Best value for your $ are the...”

Leaderboard

1
redever's picture
redever
99.2
2
Secyh62's picture
Secyh62
99.0
3
kanon's picture
kanon
99.0
4
BankonBanking's picture
BankonBanking
99.0
5
CompBanker's picture
CompBanker
98.9
6
Betsy Massar's picture
Betsy Massar
98.9
7
dosk17's picture
dosk17
98.9
8
DrApeman's picture
DrApeman
98.9
9
GameTheory's picture
GameTheory
98.9
10
Jamoldo's picture
Jamoldo
98.8
success
From 10 rejections to 1 dream investment banking internship

“... I believe it was the single biggest reason why I ended up with an offer...”