This Anthropic study examines how AI, and Claude Code in particular, is transforming the work of the company's own software engineers. It draws on a survey of 132 engineers and researchers, 53 qualitative interviews, and an analysis of internal usage data (August 2025, Claude Sonnet 4 and Opus 4 models). Its position is distinctive: observing "early adopters" within the very company developing the AI, likely representative of transformations to come elsewhere.
The figures are striking. Employees report using Claude for 60% of their work and estimate a productivity gain of roughly 50%, 2 to 3 times higher than the previous year. Most notably, 27% of Claude-assisted work simply would not have been done otherwise: more ambitious projects, "nice-to-have" dashboards, exploratory work that isn't cost-effective to do manually. Delegation nonetheless remains supervised: only 0 to 20% of work can be "fully delegated," with human validation still required for critical work.
The interviews reveal profound qualitative transformations. Engineers are developing an intuition for delegation: handing off verifiable, low-stakes, or tedious tasks first, then gradually expanding the scope — with design and "taste" remaining human, for now. Developers are becoming "full-stack," working competently in areas they would not have dared touch before. But a paradoxical concern emerges: the atrophy of the deep skills needed to write and critique code — "when producing results is so easy and fast, it becomes hard to really take the time to learn."
Social dynamics are shifting too: Claude is replacing colleagues as the first port of call for technical questions, reducing opportunities for mentorship. One senior engineer sadly reports that juniors no longer come to him. On the career front, feelings are contradictory: short-term optimism (higher-level work, managing AI systems) and long-term concern ("AI will eventually do everything and make me obsolete").
Usage data confirms growing autonomy: from 10 to 20 autonomous actions over six months, code design usage rising from 1% to 10%, and new feature implementation from 14% to 37%. 8.6% of tasks involve fixing previously deprioritized "papercuts."
Anthropic draws internal initiatives from this — new mentorship models, maintaining deep skills — and stresses the importance of actively preparing for the transition to an AI-augmented workplace, lessons that are transferable to other organizations.