ChatGPT creator probed by US regulators over AI risks

GET FREE ARTIFICIAL INTELLIGENCE UPDATES

we will send you one myFT Daily Digest Latest Email Rounding artificial intelligence News every morning.

The US Federal Trade Commission has launched a broad investigation into ChatGPT maker OpenAI, as Washington’s top antitrust regulator turns its attention to artificial intelligence.

In a letter sent to the Microsoft-backed company, the FTC asked it to share internal material on how the group uses or maintains users’ information; the data it has used to develop large language models; And the company has taken steps to address the risk of its models making “false, misleading or defamatory” statements.

The FTC declined to comment on the letter, which was first reported by The Washington Post. OpenAI did not immediately respond to a request for comment.

Generative AI products are under fire from regulators around the world, as AI experts and ethicists warn about the huge amount of personal data consumed by the technology as well as its potentially harmful outputs, ranging from misinformation to sexist and racist comments are giving ,

In May, the FTC warned the industry that it is “focusing intensely on how companies can use AI technology, including new generative AI tools, to have a real and substantial impact on consumers.” Is”.

FTC Chairwoman Leena Khan testified before the House Judiciary Committee on Thursday morning and faced strong criticism from Republican lawmakers over her tough enforcement stance.

Khan is part of a new generation of progressive antitrust officials appointed by Joe Biden’s administration seeking to crack down on anti-competitive conduct they believe has gone unchecked in the US economy for decades.

Experts are concerned about the huge amount of data being accumulated by the language model behind ChatGPT. Just two months after its launch, OpenAI had over 100 million monthly active users. Microsoft’s new Bing search engine, powered by OpenAI technology, was being used by more than 1 million people in 169 countries within two weeks of its release in January.

Users have reported that ChatGPT has fabricated names, dates and facts, as well as fake links to news websites and academic paper references, an issue known in the industry as “hallucinations”.

The FTC investigation examines technical details of how ChatGPT was designed, including the company’s work on fixing hallucinations and its oversight of human reviewers, which directly affect consumers. It has also sought information on the efforts made by the company to address consumer complaints and assess consumer understanding of the chatbot’s accuracy and reliability.

In March, Italy’s privacy watchdog temporarily banned ChatGPT while it investigated the US company’s collection of personal information following a cybersecurity breach, among other issues. It was reinstated a few weeks later, after OpenAI made its privacy policy more accessible and introduced a tool to verify the age of users.

Sam Altman, chief executive of OpenAI, has previously acknowledged that ChatGPT has vulnerabilities. “ChatGPT is incredibly limited, but good enough at a few things to create an illusory impression of greatness.” he wrote on twitter In December. “It is a mistake to depend on it for anything important right now. This is a preview in progress; We have a lot of work to do on strength and truth.”