Sources: JPMorgan Chase is giving staff the option to use its in-house LLM to write year-end performance reviews, a shortcut to the often painstaking process
Move allows employees to use bank's large language model to generate reviews from their own prompts X: @rakeshsfnyc X: Rakesh Agrawal / @rakeshsfnyc : “Write my self-review to optimize my chances of promotion. Write 360 to sink everyone else.”
Context & Ripple Effects
JPMorgan’s internal-AI push has moved from a 2023 effort to develop an investment-focused ChatGPT-like service toward operational decisions, including the later use of AI in proxy-voting work. This makes performance-review drafting a notable expansion into an employee-facing, high-volume internal process.
The move lands as professional-services firms test AI in talent workflows: McKinsey has piloted AI-assisted candidate case analysis, while banking and legal employers face concerns that AI is removing routine work that once trained junior staff.
First-order effects
- JPMorgan employees can use the bank’s LLM to turn their own prompts into year-end review drafts, reducing the writing burden in a consequential HR process.
- Managers and HR teams will receive more AI-assisted submissions, making employees’ underlying evidence and managers’ review of it more important than prose polish alone.
Second-order effects
- The option may raise expectations that other internal writing tools support sensitive people processes, while increasing pressure on employers to set clear rules for disclosure, attribution, and manager oversight.
- If AI improves the presentation of self-assessments unevenly, review processes may need more standardized evidence or calibration to prevent writing skill and prompt skill from distorting comparisons.
Third-order effects
- This points toward enterprise AI becoming embedded in systems of record for work and talent—not just optional productivity software—so governance will increasingly determine where firms permit it.
- As organizations apply AI to both recruiting and evaluation, the durable question will be whether assessment systems measure demonstrated work rather than the quality of AI-mediated narratives.
The trend: Enterprise AI is shifting from task automation into the processes that allocate opportunity, performance judgments, and organizational power.