Goldman Bans Traders And Bankers From Using AI Coding Tools
KIDieser Beitrag wurde mit Unterstützung künstlicher Intelligenz (KI) erstellt.

TL;DR

Buying for a business?Offer from Amazon

Get business pricing on office and shipping supplies

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

Goldman Sachs has restricted traders and other non-engineering employees from using advanced AI coding tools, including Claude Code and GitHub Copilot, after a period of experimentation. The bank says it is weighing productivity and risk management; details of the restriction and its duration have not been disclosed.

Goldman Sachs has restricted traders and other non-engineering employees from using advanced AI tools for coding, limiting access to engineers after a period of broader experimentation, according to Bloomberg, as reported by Finextra. The change affects staff who had used tools such as Anthropic’s Claude Code and Microsoft GitHub for work including tracking client trading flows and modelling market scenarios.

Finextra, citing Bloomberg, reported that Goldman had previously allowed bankers to experiment with coding tools such as Claude Code and GitHub but has now limited their use to engineers. Finextra’s report describes the move as a ban for traders and other employees, though the precise scope of the bank’s internal policy has not been published.

Goldman spokesperson Nick Carcaterra told Bloomberg that the bank makes selected AI tools available to employees to evaluate possible uses across the firm. He said the bank reviews factors including productivity and risk management before deciding which groups are best placed to use the tools over the longer term.

Some traders were frustrated by the restriction, according to Bloomberg’s reporting as summarized by Finextra. They had used the tools for tasks such as monitoring client trading flows and modelling market changes and positions. The report does not describe any specific incident or loss that prompted the policy change.

At a glance
reportWhen: Reported by Bloomberg and Finextra; the…
The developmentGoldman Sachs has limited advanced AI coding tools to engineers after previously allowing broader employee experimentation.

How the Restriction Changes Staff Access

The policy marks a shift from broad experimentation toward more limited access based on job role. For staff outside engineering, it could mean they can no longer use coding assistants to build or adapt internal tools for day-to-day analysis, including the trading-related tasks described in the report.

The decision also reflects a risk-management question facing large financial firms: who should be allowed to use AI tools that can generate code, and under what oversight? The bank has cited both productivity and risk in explaining its approach, but has not publicly detailed how it measured either factor or what risks it identified. The available reporting does not establish that the tools caused a security, compliance, or trading problem.

For Goldman employees, the restriction may also change how experimental tools move from an individual workflow into firm-approved software. The report says traders had been using the systems for practical work, while Goldman’s earlier public comments described non-technical staff developing tools that could later be passed to engineers. The new limit places engineers in a more central role in that process.

Products Worth Considering

Canon PIXMA PRO-200S Wireless Professional Inkjet Photo Printer

Canon PIXMA PRO-200S Wireless Professional Inkjet Photo Printer

As an affiliate, we earn on qualifying purchases.

Goldman’s Earlier AI Experiments

Goldman had previously given some bankers access to AI coding products during a period of experimentation. In comments made over the summer, Goldman CIO Marco Argenti discussed the potential of “vibe coding” by non-technical employees, with tools created in that way later passed to engineers. That account suggests the bank had been exploring ways for staff without programming backgrounds to prototype software.

The shift comes alongside public discussion within the bank about possible drawbacks. In August, Chris Churchman, a Goldman partner leading one of its flagship AI projects, warned that relying too heavily on the technology could weaken bankers’ reasoning skills. His comments were a caution about potential effects, not evidence that the current restriction was directly caused by such harm.

Goldman has also restricted access to certain AI models in a separate setting. Earlier this year, amid rising US-China tensions over AI, it was among banks that limited employee access to Claude models in Hong Kong, according to Finextra. The report does not say that the Hong Kong decision and the coding-tool restriction share the same rationale.

“We make certain AI tools available to our people to determine the best use cases across different parts of the firm.”

— Nick Carcaterra, Goldman Sachs spokesperson, speaking to Bloomberg

Policy Scope and Review Criteria

Goldman has not publicly specified when the restriction began, which exact tools or functions it covers, or whether exceptions are available for particular projects. The reporting says use has been limited to engineers, but does not provide the full text of the internal policy.

The bank also has not shared the results of its experimentation, including how it evaluated productivity, what risks it identified, or whether any measurable gains or problems drove the decision. There is no reported evidence of a specific failure, client impact, or regulatory action behind the change. It is also unclear whether the restriction is temporary, part of a broader review, or a lasting change in access.

Whether Access Will Expand Again

The next indication will be whether Goldman explains how it will assess these tools and whether non-engineering employees may regain access under defined controls. Carcaterra’s comments point to a process in which tools are tested before being assigned to groups considered suited to use them, but the bank has not announced a review timetable or a future decision date.

For now, the confirmed development is a limit on advanced AI coding tools for traders and other non-engineers, as described by Bloomberg and Finextra. Further detail about the policy, its results, and its duration remains unannounced.

Key Questions

What has Goldman Sachs restricted?

According to Bloomberg, as reported by Finextra, Goldman has limited traders and other non-engineering employees from using advanced AI coding tools, with access now restricted to engineers.

Which AI tools are mentioned in the report?

The report names Anthropic’s Claude Code and Microsoft GitHub as tools employees had previously been allowed to use. It does not provide a complete list of products covered by the new policy.

Why did Goldman make the change?

Goldman spokesperson Nick Carcaterra said the bank evaluates possible uses of AI tools, including productivity and risk management. The bank has not disclosed specific findings or said that a particular incident triggered the restriction.

Can traders still use AI at Goldman Sachs?

The report concerns advanced AI tools for coding and says their use has been limited to engineers. It does not establish whether traders can use other AI systems or whether exceptions apply.

Is the restriction permanent?

That is not clear. Goldman has not announced the policy’s duration, review schedule, or whether non-engineers could receive access again under different conditions.

Source: finextra.com

This content is for general information only and is not financial, tax or legal advice. Consult a qualified professional for decisions about your money.
You May Also Like

The Ghost Story Became a Forecast.

Thorsten Meyer analyzes Jack Clark’s recent essay revealing a bivalent forecast for AI development, with 60% chance of rapid progress by 2028 and 40% indicating fundamental limits.

The Defender’s Counter-Cascade.

On May 11, 2026, Google disclosed the first confirmed use of an AI-built zero-day exploit, highlighting the deployment gap in AI security defenses and its implications.

The Machine Economy — Capital-Heavy, Human-Light, Trading With Itself

Analysis of the emerging machine economy where AI-driven firms operate with minimal human labor, reshaping markets and economic structures.

The Anti-Mythos Weapon Of AI: Microsoft’s Signal Peak 2026 With Anthropic’s Models

Microsoft prepares to launch Project Perception, an AI security tool routing models from Microsoft, OpenAI, and Anthropic to compete with Mythos, emphasizing model routing and cost efficiency.