AI is obliterating the pipeline from finance programs to industry jobs

The vendors and the AI companies went to the firms first

The vendor map counts 39 links between 25 data vendors and 7 AI companies. Every announcement is addressed to banks, asset managers and insurers. Anthropic’s finance launch on 5 May 2026 opens with its customers: “Many leading banks, asset managers, and insurers choose Claude.” It names FactSet, S&P Capital IQ, MSCI, PitchBook, Morningstar and LSEG as connectors. It does not mention a university, a student or academic research.

The same split runs through Asia. Chinese brokers put a new model into use within hours of its release, and Japan’s regulator told firms not to hold back.

The products automate the job a graduate used to get

Anthropic shipped ten such agents. Each is a task a first-year analyst was hired to do and learned the business by doing.

The Stanford authors, Erik Brynjolfsson, Bharat Chandar and Ruyu Chen, use payroll data from ADP through June 2026. They find “no evidence of widespread, economy-wide job displacement,” they call the result descriptive, and they report that the gap for young workers “has widened steadily” since they first measured it in August 2025. The losses are concentrated where AI substitutes for the worker’s tasks.

The campus is training for the desk that is disappearing

A finance program teaches a student to pull data through a query screen, clean it by hand and write up a table. A firm now buys an agent that does all three. AACSB’s 2026 standards do not contain the word AI, so nothing requires a school to close that gap, and the standalone academic data platform is the one part of the market the new connections skip.

Mirae Asset shows the other side. The assistants who kept their careers were moved up to work that needs judgment. The firm did not hire more of them.

Learn to direct the agent before the job asks for it

• Students: finish a project with an agent doing the pull, the tests and the commits. The SEC filings guide takes an evening. Put the repository on your CV.
• Faculty: set one assignment where the deliverable is a pipeline that reruns, with its checks, and grade the checks.
• Program directors: ask the firms that hire your graduates which agents their analysts use, and give students the same ones.
• Deans: write agent access into the next data license, so students can practice on the data they will meet at work.