TL;DR
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Goldman Sachs has restricted traders and other non-engineering employees from using advanced AI coding tools after a period of experimentation. The bank said it weighed productivity and risk management; the scale of the restriction and its effect on employees’ work have not been disclosed.
Goldman Sachs has restricted traders and other non-engineering staff from using advanced AI coding tools, limiting access to engineers after a period in which the bank allowed broader experimentation. The change, reported by Bloomberg and Finextra, matters because some traders had used the tools to build or assist with internal tasks such as tracking client trading flows and modelling market positions.
Finextra reported that Goldman had allowed employees to use tools including Anthropic’s Claude Code and Microsoft GitHub tools for several months before narrowing access. Bloomberg reported the policy change; the reports did not specify when it began or how many employees are affected. Goldman has not publicly described the restriction as a permanent ban across the firm.
Nick Carcaterra, a Goldman spokesperson, told Bloomberg that the bank makes certain AI tools available to employees to test possible uses across different parts of the firm. He said the bank evaluates factors including productivity and risk management before deciding which employees are best placed to use the tools over the longer term.
According to Finextra’s account of the Bloomberg report, the change frustrated some traders who had used AI coding products for practical work, including monitoring client trading flows and modelling changes in markets and positions. The reports do not establish whether those projects have stopped, been handed to engineers, or will continue through approved processes.
Why Access Matters to Trading Teams
The restriction changes who can experiment with software that may help employees create or adapt internal tools. Traders who can no longer use the products directly may need to ask engineering teams to develop or maintain such applications, which could affect how quickly teams test ideas. That is a potential operational consequence, not a measured outcome disclosed by Goldman.
The decision also reflects a tension facing financial firms: AI coding tools may help non-specialists automate tasks, but software used in a bank can introduce risks involving reliability, data handling and oversight. Goldman’s spokesperson cited productivity and risk management as factors in the review, but the bank has not identified a particular incident or specific risk that prompted the change.
For employees and other banks watching adoption, the move suggests that access may be treated as a controlled privilege rather than a firmwide default. It does not, on the information reported, amount to a ban on all AI use at Goldman or a decision to stop using AI coding tools altogether.
Products Worth Considering
Dell XPS 13 9350 Copilot+ Laptop, 13.4-inch 120Hz Display, Intel Core Ultra 7 256V, 16GB RAM, 512GB SSD, Ubuntu Linux 24.04 LTS
Logitech MX Keys S Wireless Keyboard, Graphite
Samsung Odyssey G4 25-inch FHD Gaming Monitor
As an affiliate, we earn on qualifying purchases.
Goldman’s Earlier AI Experiments
The restriction follows a period of internal experimentation. Over the summer, Goldman chief information officer Marco Argenti praised the potential of “vibe coding” by staff without technical roles. Finextra reported that tools created through this approach were then passed to engineers, indicating a workflow in which employees could prototype ideas while technical teams handled subsequent work.
There had also been public caution within the bank about relying too heavily on the technology. In August, Chris Churchman, a Goldman partner leading one of the firm’s flagship AI projects, warned that overreliance could weaken bankers’ reasoning skills, according to Finextra. That warning is a separate concern from the productivity and risk factors cited by the spokesperson; the available reporting does not say it directly caused the new restriction.
Earlier in the year, amid rising US-China tensions over AI, Goldman was among banks that restricted employee access to Claude models in Hong Kong. That measure concerned access in a particular location and should not be conflated with the current reported limits on AI coding tools for non-engineering staff.
“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 Rationale Remain Unclear
The reports do not specify when the restriction took effect, whether it applies to every non-engineering employee or only particular tools, or whether exceptions are available. Goldman has not published a detailed policy or said whether employees can continue to use AI products for non-coding tasks.
It is also unclear what happens to projects traders began during the earlier trial period, including tools for monitoring trading flows or modelling positions. The reporting does not say whether engineers will take over those projects, whether they will be reviewed before use, or whether any have been discontinued.
Goldman has cited productivity and risk management in general terms, but the bank has not shared evaluation results, quantified productivity effects or named a security, compliance or software-quality event behind the decision. The reported frustration among some traders is attributed to Bloomberg’s reporting, not to a public statement from the affected employees.
How Goldman Sets Future Access
The next indications will be whether Goldman provides further detail about which teams can use the tools and how employees can propose new use cases. Its spokesperson’s comments point to continued evaluation, but the bank has not announced a timetable for another review or said whether access could later expand.
For now, the confirmed development is a shift toward engineer-led access following wider testing. Any assessment of its effect on trading workflows, productivity or risk controls will depend on information the bank has not yet released, including how existing employee-built projects are handled.
Key Questions
What has Goldman Sachs changed?
According to Bloomberg and Finextra, Goldman has restricted traders and other non-engineering employees from using advanced AI coding tools, limiting their use to engineers after a period of broader experimentation.
Which AI coding tools are involved?
Finextra’s report names Anthropic’s Claude Code and Microsoft GitHub tools as products employees had been allowed to use. The available reporting does not provide a complete list of affected tools.
Why did Goldman narrow access?
A Goldman spokesperson told Bloomberg the bank reviews factors including productivity and risk management before deciding who is best suited to use tools over the long term. The bank has not identified a specific incident as the cause.
Can traders still use AI at Goldman?
The reports concern advanced AI coding tools. They do not establish a firmwide ban on all AI products or clarify what non-engineering employees may use for non-coding tasks.
What happens to projects traders already built?
That has not been disclosed. The reporting does not say whether engineers will take over existing projects, whether they will be reviewed, or whether any have been stopped.
Source: finextra.com
