- Goldman is pulling back on AI coding tools. The bank has banned most staff from using advanced AI tools designed for coding, limiting Anthropic’s Claude Code and Microsoft’s GitHub to engineers only, according to internal guidance updated in October and seen by Bloomberg News.
- Some traders are frustrated. They had used the tools to monitor client trading flows, model market changes and positions, and create written and graphic content, according to a person familiar with the matter. The directive does not apply to “strats,” employees who sit at the intersection of trading, quantitative research and software engineering.
- It is a reversal from the summer. As recently as July, Chief Information Officer Marco Argenti touted “vibe coding” by non-technical employees, who built tools with AI and passed them to one of the firm’s 12,000 engineers to improve.
- Goldman framed it as a normal review. “After a period of experimentation where we assess a variety of factors including productivity and risk management, we put them in the hands of those who are best suited to leverage them for the long term,” spokesperson Nick Carcaterra said.
What Happened?
For months, Goldman Sachs allowed bankers to use Claude Code and GitHub for work before updating its internal guidance in October to restrict the tools to engineers. It is not the first time the bank has pulled back: earlier this year, it restricted the use of Anthropic’s Claude by staff in Hong Kong. Goldman has also developed its own AI assistant that combines models from OpenAI, Anthropic and Alphabet.
Why It Matters?
Wall Street is still working out how thousands of workers should use AI. Banks are spending millions of dollars on AI tools while managing concerns about cybersecurity, costs and productivity, and JPMorgan said in July it wants employees to match models to tasks to avoid using the most expensive model for something as simple as summarizing an analyst report. The move also highlights a tension for Goldman, which is benefiting from AI’s financing needs as lead underwriter on Anthropic’s initial public offering and through debt deals arranged with Nvidia. In May, John Waldron, seen as the heir apparent to CEO David Solomon, called Goldman a “human assembly line” ripe for automation.
What’s Next?
Watch whether other banks follow Goldman in narrowing access to advanced AI coding tools, and whether Goldman expands its in-house assistant as an alternative. The bank’s role on Anthropic’s IPO keeps its own use of AI in the spotlight, and investors will look for any sign that rising AI bills are changing how Wall Street firms deploy the technology.
Affected Tickers and Coins: GS, MSFT, JPM, NVDA, GOOGL
Source: Bloomberg
