AI News Daily한국어
Tuesday, November 3, 2026

Story 3 of 6

Goldman Sachs bars bankers from AI coding tools, four months after touting 'vibe coding'

First seen on X 5 hours ago@unusual_whales ♥ 3,23717 views

Goldman Sachs has restricted most staff from AI coding tools such as Claude Code and GitHub.

According to Bloomberg's Nov. 2 report as relayed by PYMNTS, Goldman had let bankers use the tools for months but limited them to engineers in internal guidance published last month. A source said some traders were frustrated; they had used AI to monitor client trading flows and create documents and graphics.

As recently as July, CIO Marco Argenti touted non-technical staff 'vibe coding' solutions that some of the firm's 12,000 engineers then refined. A spokesperson said the firm puts tools 'in the hands of those who are best suited to leverage them for the long term' after a period of experimentation weighing productivity and risk management.

The X post (unusual_whales) drew 3,200 likes. Replies mocked 'restricting AI due to productivity concerns' while others said regulated industries need a different approach.

“BREAKING: Goldman Sachs has restricted most staff from using AI tools designed for coding due to productivity and risk concerns, per Bloomberg.”

Next story →'GPT-6 Astra raised a small Qwen model's math score from 7.9% to 48.1% by editing 0.016% of its weights,' MIT says; critics say the prompt held hintsRead an AI Insight →'Prompting is over': a study path through harness, loop and graph

Get tomorrow's stories as an alert

We'll ping you when each new edition or breaking story lands. Free, no sign-up.

Daily morning AI news, new AI Insights, and a monthly briefing of top stories on the 1st.

More stories from the same day

  1. Anthropic filing says it 'may imminently meet' its threshold for AI-driven accelerated progress; Q3 revenue $17 billion, $895 billion in commitments
  2. Microsoft offers voluntary exit to about 60,000 employees as it spends $175 billion on AI infrastructure this year
  3. 'GPT-6 Astra raised a small Qwen model's math score from 7.9% to 48.1% by editing 0.016% of its weights,' MIT says; critics say the prompt held hints
  4. DeepSeek adds Creator Mode to Harness: 'tell it what you want and it writes and installs the plugin,' 1.18 million views
  5. Claude API 'message threads' in public beta: send only new messages and the server remembers the conversation
See all of Tuesday, November 3, 2026 →

* Summaries and publishing on this page are done by an automated agent. Check the facts against the original links.

Privacy PolicySupportSend feedbackDiscord botClaude Code & CodexRSSListed on Sell With boost