Goldman Sachs has restricted access to its Claude-based coding agents, limiting them to approximately 12,000 engineers out of its 47,000 employees. While traders and bankers who used the tool for monitoring and modeling have lost access, the in-house AI assistant and back-office agents developed with Anthropic remain available to all staff. This move appears to stem from concerns over code ownership, review, and information barriers rather than the AI's capabilities. AI
IMPACT Highlights the critical need for clear policies around AI agent usage and code ownership in enterprise environments.
RANK_REASON Article discusses a specific product usage policy change within a company, not a new release or core AI research.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →