June 7, 2023

Main corporations, together with Amazon, Goldman Sachs, and Verizon, have banned or restricted the OpenAI-owned chatbot, which exploded in reputation late final yr. Many have cited privateness dangers for the clampdowns since ChatGPT makes use of knowledge from conversations to enhance its accuracy, though customers can decide out through ChatGPT’s settings or a Google kind.

After all, a number of corporations use ChatGPT and different generative A.I. instruments like Google Bard to enhance productiveness and release time spent on menial duties like drafting emails or reviewing code. Corporations like Coca-Cola and Bain & Firm have even inked partnerships with OpenAI.

However for a lot of employers, the safety dangers far outweigh the productiveness profit, as evidenced by Samsung. In early April, the corporate reported that workers by accident leaked confidential inside supply code and assembly recordings whereas utilizing the chatbot (Samsung has since banned ChatGPT’s use for workers). Apple turned the newest firm to ban ChatGPT, on the time of publication, citing related issues over knowledge leaks.

These corporations usually are not alone of their apprehension. Italy quickly banned the service, stating that OpenAI isn’t compliant with Europe’s GDPR privateness legislation, and the E.U. is near passing a groundbreaking legislation governing A.I.’s use within the bloc.

Whereas the chatbot is at the moment restricted in sure workplaces, which may not all the time be the case. Some corporations, like JPMorgan Chase and Deutsche Financial institution, have indicated they’ll permit ChatGPT if it’s absolutely vetted for secure use. Others, like Goldman Sachs and Samsung, are growing their very own A.I. instruments for worker utilization.

Under are corporations which have banned or restricted workers from utilizing ChatGPT at work.


Apple is proscribing employees from utilizing ChatGPT and different third-party A.I. instruments, citing issues about confidential knowledge leaks. The tech large additionally informed workers to not use the automated software program code-writing program Copilot, developed by GitHub and OpenAI (each owned or backed by Microsoft). Apple, which tapped Google veteran John Giannandrea to steer its A.I. efforts in 2018, is reportedly growing its personal A.I. instruments.

Financial institution of America

Financial institution of America added ChatGPT to its checklist of unauthorized apps that workers are prohibited from utilizing for enterprise, insiders informed Bloomberg in February. The financial institution is considered one of many to implement stricter compliance measures round inside communications after U.S. regulators levied over $2 billion in fines for the agency’s failure to watch worker use of unauthorized messaging apps like WhatsApp.


CEO Michael Weening introduced on LinkedIn that the Calif.-based telecommunications firm banned ChatGPT throughout all enterprise capabilities and units in April. He cited Samsung’s current leak because the driving motive for the ban, writing it’s “disconcerting” that ChatGPT may expose delicate info like confidential inside memos or buyer contracts below NDA to outsiders. “The world is altering. Large time. Individuals must learn the tremendous print (or lack of it) and educate groups.”


Citigroup added ChatGPT to its normal firm-wide controls for third-party software program, the place sure classes of internet sites are robotically restricted. “We’re actively exploring numerous avenues that might give us a greater understanding of the advantages and potential related dangers of utilizing this expertise,” a spokesperson tells Fortune.

Deutsche Financial institution

Entry to ChatGPT has been disabled for Deutsche Financial institution employees since a minimum of February. The restriction is normal observe for third-party web sites, a spokesperson tells Fortune, noting its intention is to guard the financial institution from knowledge leakage reasonably than a sign of the device’s usefulness. Within the meantime, the financial institution says it would consider the best way to finest use the platform whereas defending its personal and shopper’s knowledge. The financial institution is growing A.I. chatbots and was an early enterprise companion of Google Cloud’s generative A.I. choices.

Goldman Sachs

Like Citigroup, Goldman Sachs blocked entry to ChatGPT for workers by an automated restriction on third-party software program, as Bloomberg first reported in February. 

However Goldman is within the technique of growing its personal generative A.I. instruments. Goldman chief info officer Marco Argenti mentioned the financial institution has a number of “proof of ideas” to streamline duties like doc classification and categorization, summarizing earnings calls, or analysis for a each day digest.

JPMorgan Chase

America’s largest financial institution restricted using ChatGPT for workers in late February, an individual accustomed to the choice informed Bloomberg. The choice is a part of normal controls for third-party software program, the individual mentioned. However the financial institution could also be open to utilizing the device sooner or later, noting in its 2022 annual report that it’s “imagining new methods to reinforce and empower workers with A.I. by human-centered collaborative instruments and workflow, leveraging instruments like massive language fashions, together with ChatGPT.” The financial institution didn’t reply to Fortune’s request for clarification on the way it’s restricted. 

Northrup Grumman

The aerospace and protection expertise firm blocked the device earlier this yr, claiming it doesn’t share firm or buyer info with outdoors instruments till vetted. One worker informed the Wall Avenue Journal he used the chatbot for months previous to the ban, asking it idea questions on code earlier than turning the plain language response into code itself. He in contrast the bot to a “actually cool, affected person mentor who’s not going to be irritated by you asking numerous questions.” Northrop Grumman didn’t reply to Fortune’s request for remark.


The telecommunications large can be involved about privateness and safety with ChatGPT. Verizon knowledgeable workers in mid-February that this system just isn’t accessible through company methods as a result of danger of shedding management of delicate info like buyer knowledge and supply code. “Our high precedence is our 4 stakeholders: communities, prospects, buyers/shareholders, and society, and we’ve got to be considerate when introducing a brand new and rising expertise corresponding to a ChatGPT,” a communications supervisor at Verizon wrote to workers.


South Korea-based Samsung banned worker use of ChatGPT and different generative A.I. instruments in early Could. The restriction got here after engineers by accident leaked confidential info, together with inside code and assembly recordings, in April, by importing it to the chatbot. Staff are banned from accessing the instruments on company-owned units and its inside networks. Nevertheless, Samsung creating its personal A.I. instruments that workers can use for software program improvement, translation, and summarizing paperwork.

Under are corporations that haven’t outright banned ChatGPT however requested workers to not share confidential info on the platform:


Accenture restricts its 700,000 workers from utilizing ChatGPT and different generative A.I. instruments whereas coding, noting they need to get hold of permission to share firm or shopper knowledge. Accenture didn’t reply to Fortune’s request for clarification on the restrictions.


Amazon warned workers towards sharing confidential firm info with ChatGPT, citing cases during which the chatbot’s responses appeared just like inside knowledge, in accordance with messages considered by Insider. 

Amazon workers have reported receiving a popup warning that they’re coming into a third-party web site that “is probably not authorised to be used by Amazon Safety” when visiting ChatGPT’s web site. Engineers searching for coding assist are inspired to make use of an inside A.I. device known as CodeWhisperer. However workers usually are not outright restricted from utilizing ChatGPT or related instruments. “We now have safeguards in place for worker use of those applied sciences, together with steering on accessing third-party generative A.I. providers and defending confidential info,” a spokesperson tells Fortune.

PwC Australia

The Australian division of the skilled providers agency warned workers towards utilizing any materials or info outputted by ChatGPT and prohibited sharing agency or shopper knowledge with the device, in accordance to an inside publication. Nevertheless, it nonetheless permits entry to ChatGPT and encourages workers to experiment personally. Jacqui Visch, PwC Australia’s chief digital and knowledge officer, informed the Australian Monetary Assessment the corporate is exploring cybersecurity and authorized concerns earlier than allowing the chatbot for any enterprise use.