Goldman Sachs bans non-engineers from using advanced AI tools
Goldman Sachs has banned traders, bankers, and other staff from using advanced AI coding tools because of risk and productivity concerns. The bank is limiting access to tools like Anthropic’s Claude Code and Microsoft’s GitHub to engineers and so-called “strats,” an investment banking term for workers who specialize in quantitative engineering and analytics, Bloomberg reported this week. The move…
Goldman Sachs has restricted access to advanced AI coding tools like Anthropic’s Claude Code and Microsoft’s GitHub to engineers and investment banking specialists known as "strats." The bank cited risk and productivity concerns as the primary reasons behind the ban, affecting traders and other staff members who previously utilized the technology for data monitoring, financial modeling, and graphic creation.
This decision has reportedly irritated some traders who relied on these tools for their daily tasks. The move contradicts comments made by Goldman's chief information officer, Marco Argenti, in July, where he applauded non-technical Goldman staff for using AI to create software prototypes, which were then refined by engineers. A Goldman spokesperson clarified that the firm permits staff to experiment with certain AI tools to identify the most suitable applications across the organization, taking into account factors such as productivity and risk management.
Once the experimental phase concludes, the company plans to grant access to these tools to individuals best equipped to leverage them long-term. The decision to limit AI tool access aligns with a broader trend among Wall Street institutions grappling with how to balance speed and safety in AI adoption. JPMorgan Chase's Chief Information Officer, Lori Beer, expressed the challenge of managing risks in real-time as AI technology evolves rapidly.
"We want to move at speed, but sometimes we must pull back to get controls in place," she noted, emphasizing that the technology is outpacing the capabilities to manage it effectively. Bank of America CEO Brian Moynihan echoed similar sentiments, urging caution regarding the use of AI agents without human supervision, particularly in creating pitchbooks and spreadsheets.
He stressed the importance of human oversight to ensure the accuracy of AI-generated data, as banks are keen on capitalizing on the benefits that AI can bring. Despite the restrictions, Goldman Sachs continues to invest in AI initiatives, including a proprietary chatbot accessible to all employees and a coding assistant for developers.
The bank is also piloting autonomous coding agents and AI tools designed to assist investment bankers in compiling research.
Written by urgent.news from Fast Company's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.