Many US workers are turning to ChatGPT to help them with basic tasks, a Reuters/Ipsos survey found, despite concerns that have prompted employers such as Microsoft and Google to curtail its use, Reuters and News.ro reported.

Businesses around the world are considering how best to use ChatGPT, a chatbot application that uses generative artificial intelligence to communicate with users and answer countless queries.

However, security firms and companies have expressed concern that this could lead to leakage of intellectual property and strategies.

Examples of people using ChatGPT to help with their daily work include writing emails, summarizing documents, and doing preliminary research.

About 28 percent of respondents to an online artificial intelligence (AI) survey conducted July 11-17 said they regularly use ChatGPT at work, while only 22 percent said their employers explicitly allowed such external tools.

The Reuters/Ipsos poll of 2,625 American adults had a confidence interval, a measure of precision, of about 2 percentage points.

About 10 percent of respondents said their bosses had outright banned external AI tools, and about 25 percent didn’t know if their company allowed the technology.

After launching in November, ChatGPT became the fastest growing app in history. This has caused excitement and alarm, putting the OpenAI developer at loggerheads with regulators, particularly in Europe, where the company’s massive data collection has drawn criticism from privacy watchdogs.

Evaluators from other companies can read any generated chats, and researchers have found that an AI-like AI can reproduce the data it absorbed during training, posing a potential risk to sensitive information.

“People don’t understand how data is used when they use generative artificial intelligence services. This is critical for companies because users do not have a contract with many AIs (as they are free services), so companies would not take the risk through the normal evaluation process,” said Ben King, VP of Customer Trust. at the corporate security firm Okta.

OpenAI declined to comment when asked about the implications for individual employees using ChatGPT, but pointed to a recent company blog post that assured enterprise partners that their data would not be used to train the chatbot going forward unless they gave express permission .

When people use Google Bard, it collects data such as text, location and other usage information. The company allows users to delete past actions from their accounts and request removal of content entered into AI.

Google declined to comment when asked for more details. Microsoft did not immediately respond to a request for comment.

An employee of the US company Tinder said that employees of the dating app used ChatGPT for “innocuous tasks” such as writing emails, although the company does not officially allow this.

“These are normal emails. Very low-key, like creating funny calendar invitations for team events, farewell emails when someone leaves… We also use it for general research,” said the employee, who declined to be identified because he was not authorized to speak to a reporter.

The employee said Tinder doesn’t have a “no ChatGPT” policy, but employees still use it “in a general way that doesn’t reveal anything about us being on Tinder.”

Reuters could not independently confirm how Tinder employees use ChatGPT.

Tinder said it provides “regular guidance to employees on security and data best practices.”

In May, Samsung Electronics banned employees worldwide from using ChatGPT and similar AI tools after an employee was found to have uploaded sensitive code to the platform.

“We are studying measures to create a safe environment for the generative use of artificial intelligence, which increases the productivity and efficiency of employees. However, until these measures are ready, we are temporarily restricting the use of generative artificial intelligence through the company’s devices,” Samsung said in a statement on August 3.

Reuters reported in June that Alphabet warned its employees about their use of chatbots, including Google’s Bard, as it sells the app around the world.

Google says that while Bard may provide unsolicited code suggestions, it helps programmers. He also said he is committed to being transparent about the limitations of his technology.

Some companies told Reuters they are implementing ChatGPT and similar platforms while maintaining security.

“We’ve started testing and learning how AI can improve operational efficiency,” said a Coca-Cola spokesman in Atlanta, Georgia, adding that the data remains behind a firewall.

“Internally, we recently launched our enterprise version of Coca-Cola ChatGPT to improve productivity,” the spokesperson said, adding that Coca-Cola plans to use AI to improve the efficiency and productivity of its teams.

Meanwhile, Don Allen, Tate & Lyle’s chief financial officer, told Reuters the global ingredients maker was testing ChatGPT “to find a way to use it in a secure way”.

“We have different teams deciding how they want to use it through a series of experiments. Should it be used in investor relations? Should we use it in knowledge management? How can we use it to accomplish tasks more effectively?”

Some employees say they cannot access the platform at all from their corporate computers.

“It’s completely off limits at the network office, like it’s not working,” said a Procter & Gamble employee, who spoke on condition of anonymity because he was not authorized to speak to the media.

P&G declined to comment.

Reuters could not independently confirm whether P&G employees could not use ChatGPT.

Paul Lewis, chief information security officer at cyber security firm Nominet, said companies are right to be cautious.

“Everyone understands the benefits of this advanced capability, but information is not completely secure and can be manipulated,” he said, referring to the “malicious means” that can be used to force AI chatbots to reveal information.