Kerry-Anne Mathieson
April 2, 2026
JPMorgan received 630,000 internship applications in 2025. Only 4,100 students got in. Here’s exactly how AI is reshaping consulting, banking, and data science — and what you can do right now to be one of them.
Let’s start with the number that should change how you think about the next three years of your academic life. In 2025, JPMorgan received 630,000 applications for its summer internship programs — up 25% from the year before. The firm hired roughly 4,100 interns. That is a 0.6% acceptance rate, making it statistically harder to get into a JPMorgan summer program than it is to get into Harvard.
Here is what that number doesn’t tell you: the 4,500 interns JPMorgan hired the year before that. The trend is fewer seats, more competition, and a fundamentally different job description for the students who do get in.
AI is not the only force reshaping early-career finance and consulting. But it is the one that will accelerate everything else — and the one that almost no career resource is giving you a straight answer on.
This piece will.
Before we talk about strategy, let’s be specific about what is being automated — because vague warnings about “AI disruption” are not useful. Here is what AI tools can already do at human-level speed or better in finance and consulting contexts:
Research analysts at bulge-bracket banks historically spend 60–70% of their time on data collection, model building, and synthesis — tasks that AI now handles at a fraction of the time. Deloitte projects that generative AI could generate an additional $3.5 million in revenue per front-office employee by 2026.
A 2024 survey of hiring managers found that 70% believe AI can do the work of interns, and 57% said they trust AI’s output more than the work of recent graduates. That is a striking number — and it explains why entry-level tech hiring fell 25% year-over-year in 2024.
“First drafts that used to take days now take minutes. But strategic thinking, client customisation, and deal structuring still require human expertise.”
The key phrase there is “still require.” Because the list of things AI cannot do — at least not reliably, not at the level clients actually trust — is exactly the list of skills you should be building right now.
Management consulting, investment banking, and data science are all being transformed — but in meaningfully different ways. Here is an honest breakdown of each.
Despite automation trends, top consulting firms continue to recruit summer interns, and the big names — McKinsey, Bain, BCG, Deloitte, EY, and PwC — are all running full programs. But the recruiting process itself is changing in ways that signal exactly what these firms are looking for.
In late 2025, McKinsey introduced an AI-driven interview component to its final-round assessment for Business Analyst candidates. The pilot requires candidates to collaborate with Lilli — McKinsey’s proprietary AI platform — as part of their evaluation. As of early 2025, Lilli processes more than 500,000 prompts per month internally. It’s not a test of technical AI knowledge. It’s a test of whether you can work with AI the way a consultant actually does: evaluate its outputs, direct it with clarity, and apply judgment to what it produces.
This isn’t theoretical. It’s already being tested on the students recruiting right now, and it will be standard practice across MBB recruiting cycles by summer 2026.
The recruiting timeline has also become dramatically more compressed. McKinsey’s Summer 2027 Business Analyst Intern applications opened January 1, 2026, with a deadline of March 29 — nearly four months earlier than the prior year’s cycle. BCG and Bain are on similar timelines. If you’re not thinking about this before sophomore year, you’re already behind the students who will get the offers you want.
The investment banking picture is the most urgent of the three tracks — because there is a specific technology deployment timeline that should inform when and how you recruit.
OpenAI’s banking-focused AI system — widely referred to as Mercury in industry reporting — is expected to be live at bulge-bracket banks by 2027. According to analysis from Prospect Rock Partners, OpenAI hired over 100 former investment bankers specifically to develop tools that automate the core analytical and modelling work currently done by junior analysts. The timeline matters enormously depending on your graduation year.
Students recruiting for 2026 positions will likely arrive before full deployment, giving them 6–12 months of relatively traditional training before the analyst experience changes significantly. Students recruiting for 2027 will arrive into a different world. Industry analysis from Mergers & Inquisitions projects a flattening of the traditional analyst-to-MD hierarchy, with fewer junior seats but higher expectations for the analysts who remain.
A practical question to ask in every banking interview: “When do you expect to deploy AI tools for analyst-level work? What percentage of the analyst role do you estimate AI will handle by the time I’m a second-year? How is your training programme adapting for that?” Banks that give substantive answers are the ones worth joining.
One more number that should recalibrate your assumptions: the $200K+ analyst compensation that many students are factoring into their decision to pursue banking may not exist by 2027 in its current form.
Don’t make a career decision based on 2024 compensation data without verifying what 2027 offers actually look like.
Data science gets a different kind of disruption. The headline risk — “AI will replace data scientists” — is simultaneously overstated and understated depending on which skills you’re talking about.
Routine model building, pipeline creation, and exploratory data analysis are increasingly handled by AI assistants. A student who can write Python and build a logistic regression model is no longer differentiated. Recent Duke University and Federal Reserve research notes that AI’s productivity impact is strongest in high-skill services like finance — precisely where data science sits.
What is becoming more valuable: the ability to define what problem to model in the first place, to govern models and take accountability when they fail, to translate technical outputs into decisions that non-technical stakeholders will act on, and to understand the ethical and regulatory risks of AI outputs in financial contexts. These are not Python skills. They are judgment, communication, and domain expertise.
A data science student who pairs strong technical foundations with genuine business communication skills and an understanding of AI governance is building a durable career. A student who treats data science as a technical credential alone is building on eroding ground.
This is the most practically useful frame for students currently in their freshman or sophomore year at a US university.
The good news: the preparation that wins in this environment is more interesting and more learnable than the preparation that used to win. You don’t need to become an AI engineer. You need to become someone who works confidently and critically alongside AI — in a professional context, on real problems, with real stakes.
This is not about taking an AI course. It’s about learning to use AI tools the way a junior analyst at Goldman or a first-year at McKinsey actually uses them: to generate a first draft of an analysis, to pressure-test a financial model, to synthesise research from multiple sources into a coherent memo.
The McKinsey AI interview pilot is the clearest signal available: firms are testing whether you can collaborate with AI productively in a professional context.
The shift across all three career tracks is from execution to oversight and interpretation. The skills that become more valuable as AI automates execution are:
These are not soft skills in the dismissive sense. They are the specific capabilities that firms are explicitly naming as their new hiring criteria — and they cannot be acquired through coursework alone.
A GPA and a course list are not differentiated anymore, and they haven’t been for several years. What firms are looking for is evidence of applied work on real problems, in contexts where the stakes were real and the outputs mattered.
International exposure matters more than it used to, not as a line on a resume but because it develops the adaptability and cross-cultural communication that AI cannot provide. The ability to navigate ambiguity in an unfamiliar environment — to build relationships, solve problems without a playbook, and communicate across difference — is precisely what firms are paying a premium for as AI handles the structured work.
This is the core of what iXperience programs are designed to build: not just technical skills, but the kind of applied professional experience in real international contexts that creates genuinely differentiated candidates. Our alumni who go on to Goldman, McKinsey, and equivalent firms consistently point to the applied project work — working on real business problems with real companies — as the experience that made their applications different.
The most underutilised tool in recruiting is the question. Students who ask the right questions in interviews signal exactly the kind of strategic maturity and AI awareness that firms are looking for. Three questions worth having ready for any banking or consulting recruiter:
If you’re a freshman or sophomore at a US university with your eye on a top summer analyst program, the practical checklist is straightforward:
The students who will get the seats that are genuinely worth having in 2026 and 2027 are not the ones who were the most anxious about AI — they’re the ones who got curious about it early, built real professional experience alongside it, and arrived at recruiting season with a profile that the 0.6% who get in actually have.
That profile is buildable. The question is whether you start building it now.
Sources
Financial News London — Record applications to Wall Street internship programs, 2025
Stack Overflow — AI vs Gen Z: How AI has changed the career pathway for junior developers, 2025
Integrity Research — The AI Revolution in Investment Research, 2026
Cyndx / Deloitte — The Future of the Investment Banking Analyst: Augmented, Not Replaced
Management Consulted — McKinsey AI Interview Now a Part of Final Round, January 2026
Prospect Rock Partners — OpenAI Mercury banking deployment timeline, 2025
Mergers & Inquisitions — AI for Financial Modeling: Will It Kill Investment Banking?
Fortune / Duke University & Federal Reserve — CFOs believe AI is paying off. Researchers aren’t so sure—yet, March 2026
Extern — Consulting Internships Summer 2027: Firms & Timeline
CaseLane — Consulting Application Deadlines 2025–26
Join thousands of students who've transformed their futures in just 6 weeks. Spots are limited by course and location.
Transforming students into industry-ready professionals through immersive learning and real-world experience.
4,500+ Alumni
85% at Fortune 500's
80+ Global internship partners
Partnered with
Our programs.
Our programs.
Partners.
