Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Goldman Sachs has just launched his AI generative assistant throughout the company, which makes it at the disposal of all employees in what the bank calls an important step in its technological strategy.
This decision follows more than a year of internal development and tests involving more than 10,000 employees piloting the tool. The GS AI assistant is an interface of conversational AI which allows employees to interact safely with large models of language such as GPT and Gemini, with firewall as part of secure Goldman compliance.
“We have been developing AI applications and infrastructure and automatic learning for several years, including several generative tools fed by AI that transforms our way of working,” said information director Marco Argenti in a memo seen by Gizmodo.
These tools include a developer co -pilot for coding, a translation tool for internal teams and an emerging “banking co -pilot” designed to rationalize workflows for investment bankers. The GS AI assistant, however, is the first generative AI system to be deployed throughout the business. The official objective is to help employees manage tasks such as the summary of complex documents, content writing and data analysis, work that can consume hours of human time.
The initiative does not consist in replacing jobs but improving the operation of employees, said a person familiar with the launch in Gizmodo. Goldman Sachs, said the source, hopes that the tool will allow its employees to stimulate efficiency.
But this decision is part of a quiet arms race at Wall Street, where companies like Citi, Bank of America and Morgan Stanley also deploy AI chatbots to automate tedious and white work that has been using legions of junior bankers for decades.
According to experts, AI already transforms these banks. Instead of deploying analyst armies to manually scan legal documents, for example, some on Wall Street now use AI to identify key clauses in contracts and flag elements that require human attention.
Some banks have even built AI to manage margin calls. “When a customer responds to a margin call email with` `yes ”,” no “or a vague question, the AI analyzes the free text answer and decides what to do,” said a banker of a large investment bank in Gizmodo. If the AI is sufficiently confident, the system automatically reserves the call. No human needed.
Even management tasks are automated. The banker said that their company uses AI to help managers generate staff exams and objectives, freeing time and ensuring that the documentation is more polite.
Although the official line is that the AI frees employees for “higher value work”, the real consequence is a reduced need for human work. The banker confirmed that, because their AI system now deals with 85% of all customers’ responses for margin calls, “the operations team has avoided hiring 30 new people.”
At Goldman Sachs, the deployment of AI started last year with a developer co -pilot which is now used by more than 12,000 engineers, producing substantial productivity improvements. After this success, the company began to extend the GS AI assistant, with an extremely positive internal feedback, which prompted this week’s business scale.
“Our residents of the company already integrate the generator in their workflows, stimulate productivity gains for our teams and offer advantages to our customers,” wrote Argenti in internal memo.
Although the use of the assistant is optional, the message is clear: everyone is encouraged to try it. For the technological industry, Goldman’s decision is a major validation, giving legitimacy to the generative role of AI in financial services. It also reflects a broader trend: with AI being cooked in software such as Microsoft and Outlook teams, employees now use it by default.
Of course, this productivity revolution has a human cost. If an AI tool replaces the need for 30 back office employees in a corner of a bank, what happens when the whole industry increases this?
