JPMorgan Starts Factoring ChatGPT and Claude Use Into Engineer Evaluations

JPMorgan is making AI use part of everyday work rather than an optional habit. The bank is asking about 65,000 engineers and technology specialists to use AI tools regularly, and managers are already tracking how actively employees do so.

A new working standard

This is not a one-off initiative for experimentation. It is an attempt to build generative AI into the ordinary production process. In effect, JPMorgan is telling its technical teams that working without such assistants is no longer a neutral choice.

Previously, an employee could try ChatGPT or Claude on their own initiative. Now, using such services is becoming an expected part of the daily routine.

For a large bank, this is a significant shift. In the financial sector, new tools usually go through a long approval cycle because of security, compliance and output-quality requirements. So the fact that one of the world's largest financial institutions is moving AI from the "you can try it" category to the category of a work norm shows that corporate adoption has entered its next stage, in which what is valued is no longer interest in the technology but discipline in using it.

How it will be measured

According to available information, managers track how often employees use AI tools, and that activity may affect performance reviews.

This changes the logic of adoption itself. When a company simply opens access to a new technology, usage grows slowly and unevenly. Once metrics and a link to performance evaluation appear, AI starts to be seen as part of professional effectiveness rather than the personal choice of an individual engineer.

In practice, this approach may push teams to use AI for a wide range of tasks:

• drafting code and documentation
• quickly finding options for solving technical problems
• summarizing long specifications and internal materials
• speeding up review, analysis and communication between teams
• preparing first versions of texts, queries and work notes

It also matters that, judging by the description, the bank is promoting not a single closed tool but products familiar to the market, such as ChatGPT and Claude. That makes the signal stronger: JPMorgan is not just testing trendy software but building a habit among employees of working alongside several AI assistants, and doing so regularly.

Why the culture is changing

The main news here is not the tracking itself but the change in the corporate standard. Not long ago, companies were debating whether generative AI should be allowed at all inside sensitive processes. Now the question is shifting: why isn't an employee using AI if they have access to tools that promise to speed up their work?

For engineers, this means a new unspoken expectation: the ability to work with AI is becoming part of basic digital literacy, almost on a par with proficiency in IDEs, version control systems and internal knowledge bases.

The approach has a downside. If frequency of use starts to influence an employee's evaluation, companies will have to carefully separate real productivity from formal "click-padding" in the interface. Otherwise the metric easily turns into a game of query volume rather than result quality. Even with that risk, though, JPMorgan's decision shows where the market is heading: AI is increasingly seen not as a bonus for enthusiasts but as a mandatory layer of everyday office and engineering work.

What this means

JPMorgan is effectively showing a future model for large employers: access to AI alone is no longer enough, and measurable use is becoming what matters. If this approach takes hold, then in the coming years AI skills will be assessed as pragmatically as development speed, documentation quality and contribution to team results.