Skip to main content
Goldman Sachs employees, traders, and bankers, are restricted from advanced AI coding tools, impacting their workflow.

Editorial illustration for Goldman Sachs cuts traders and bankers off from advanced AI coding tools

Goldman Sachs Restricts AI Coding Tools to Engineers

Goldman Sachs cuts traders and bankers off from advanced AI coding tools

• 4 min read

Goldman Sachs has locked most of its staff out of advanced AI coding tools, including Anthropic's Claude Code and Microsoft's GitHub, according to internal guidelines from October reported by Bloomberg. Access now belongs almost exclusively to engineers. Traders and bankers, who had spent months using these tools to track trading flows and draft content, are no longer part of the plan.

The shift marks a sharp reversal from where Goldman stood just a few months earlier. In July, CIO Marco Argenti talked up "vibe coding" among non-technical staff, describing a workflow where a trader or someone in asset management could sketch an idea, build a rough prototype with AI, then pass it to an engineer for production. That open-door approach has now closed, and Bloomberg reports the change has stirred frustration among employees who'd gotten used to the access.

Goldman spokesperson Nick Carcaterra pointed to a mix of concerns behind the decision, including productivity and risk management, without laying out specifics. The framing from the bank is that the experimental stage has run its course. What that means for the thousands of traders and bankers who'd built these tools into their daily routines is the real question now.

Goldman Sachs has cut off most of its employees from advanced AI coding tools like Anthropic's Claude Code and Microsoft's GitHub. Only engineers can still use them, according to Bloomberg, citing internal guidelines from October. Traders and bankers had spent months using the tools to monitor trading flows and create content.

Why this matters

Goldman's reversal is a useful data point for anyone selling "AI for everyone" inside a regulated enterprise. Marco Argenti talked up these tools in July; four months later, compliance and risk teams apparently won the argument, and access got narrowed back to engineers. That gap between CIO enthusiasm and what actually survives an internal audit is the real story here.

For founders pitching coding agents to banks, insurers, or anyone with a legal and compliance department, the lesson isn't that traders can't use Claude Code, it's that permissioning, auditability, and code-review trails matter more than raw capability once a tool leaves the sandbox. The frustration Bloomberg describes among Goldman's traders and bankers also tells us something: people got used to having these tools fast, and clawing back access after months of use is its own management problem. Watch whether other banks follow Goldman's lead or hold firm, and whether Anthropic or Microsoft respond with enterprise features built specifically for this kind of rollback risk.

Common Questions Answered

Why did Goldman Sachs restrict access to advanced AI coding tools like Claude Code and GitHub?

According to internal guidelines from October, Goldman Sachs limited access to these tools primarily due to compliance and risk concerns. The company's compliance and risk teams apparently won the internal argument against broader employee access, resulting in a significant reversal from the CIO's earlier enthusiasm for the tools in July.

Which employees at Goldman Sachs still have access to advanced AI coding tools?

Only engineers at Goldman Sachs retain access to advanced AI coding tools like Anthropic's Claude Code and Microsoft's GitHub. Traders and bankers, who had previously spent months using these tools to track trading flows and draft content, are now locked out of the platform.

What specific use cases were traders and bankers using AI coding tools for at Goldman Sachs?

Traders and bankers at Goldman Sachs had been using advanced AI coding tools to monitor and track trading flows as well as to create and draft content. These tools had been part of their workflow for several months before access was restricted in October.

What does Goldman Sachs' decision reveal about AI adoption in regulated financial institutions?

Goldman's reversal demonstrates a significant gap between initial CIO enthusiasm for AI tools and what actually survives internal compliance and risk audits at regulated enterprises. For founders pitching coding agents to banks and other regulated companies, this illustrates that legal and compliance departments often have the final say on technology adoption, regardless of initial executive support.

LIVE22:29SpaceXSI cuts Grok Bot quotas weeks after Musk called it fastest-growing