AI News
All stories
Reported

Goldman Sachs limits Claude Code and GitHub agents to its engineers

New internal guidance at Goldman Sachs returns AI coding agents to engineers, weeks before the bank takes Anthropic public.

AI in Practice · · 2 min read

Most traders and bankers at Goldman Sachs can no longer use AI coding tools at work. Internal guidance updated in October limits Anthropic's Claude Code and Microsoft's GitHub to the bank's engineers, Bloomberg News reported. The change followed a review of productivity and risk concerns. It lands weeks before Goldman takes Anthropic public, and it comes a few months after the bank's own chief information officer publicly praised traders' AI-built tools.

As recently as July, Chief Information Officer Marco Argenti was promoting what he called "vibe coding" by Goldman's non-technical staff. He told Bloomberg that traders and bankers had built tools with AI and passed them to the firm's engineers, of whom there are about 12,000, to be improved. The tools at issue are agents that can act on their own. Claude Code reads files, writes programs and runs them, typically with the same access as the person using it. Code an engineer writes with it usually passes through review, testing and deployment controls before it reaches anything important. A script a trader builds to watch client flows can end up in daily use without passing through any of those. Banks are also required to keep information barriers between groups, such as bankers working on deals and traders who could profit from knowing about them. Goldman built its internal AI platform around controls like these, and a tool a trader assembles with an agent was not designed with them in mind. Engineers can only check such a tool if someone sends it their way. The affected traders had used the tools to monitor client trading flows, model market moves and positions, and put together written and graphic material. Bloomberg reported that the change frustrated some of them. Goldman spokesperson Nick Carcaterra said the bank makes certain AI tools available "to determine the best use cases" across the firm, and that after experimentation the bank puts tools "in the hands of those who are best suited to leverage them for the long term." Bloomberg's sources said productivity was part of the review.

The restriction does not cover strats, employees who work across trading, quantitative research and software engineering. This is not Goldman's first pullback. In April the bank removed Claude from its internal AI platform for bankers in Hong Kong. The Financial Times reported that Goldman had read its contract with Anthropic strictly, and an Anthropic spokesperson said Claude had never been officially supported in Hong Kong. The Hong Kong restriction rested on location and contract terms. The new one rests on role. Goldman's in-house AI assistant, which combines models from several companies, remains available across the firm. The restriction applies to third-party agents that can write and run code by themselves.

Goldman is still working closely with Anthropic. Earlier this year, Goldman said Anthropic engineers had spent six months at the bank helping build Claude-based agents for trade accounting and client onboarding. Those agents sit in the back office under Goldman's engineering teams. Goldman and Morgan Stanley are also lead underwriters on Anthropic's IPO, according to the S-1 filed October 26. Anthropic is expected to list on Nasdaq under the ticker ANT before the end of November. Goldman's bankers will be marketing shares in a company whose coding agent most of them can no longer use.

What changed

Most Goldman Sachs traders and bankers can no longer use Claude Code and GitHub coding agents at work; access is now limited to engineers.

Type: Breaking change · Applies to: Claude Code, GitHub

Why it matters

Firms that let non-engineering staff build with coding agents may need to reassign ownership, oversight and access controls as those tools move into regulated business functions.

What you can do

Review which non-engineering teams run coding agents and confirm who owns, supervises and approves their use.

Original source

We have no link to the original source for this story yet.

Further reporting

Where we saw it

1 source

State of AI (Melissa Rosenthal)

AI wrote this summary from the sources listed, and nobody reviewed it individually before publication. Check the original source before you rely on it. How it works