The U.S. Federal Trade Commission has launched an investigation into OpenAI, the artificial intelligence startup behind ChatGPT, over whether chatbots harmed consumers by collecting data and publishing false information about individuals.
In a 20-page letter sent to the San Francisco-based company this week, the company said it was also investigating OpenAI’s security practices. In the letter, the FTC asked the startup dozens of questions, including how it trains its AI models and what it does with personal data.
The contents of the survey are previously reported The Washington Post reported that a person familiar with the investigation confirmed.
The FTC investigation represents OpenAI’s first major regulatory threat. The company’s co-founder Sam Altman testified before Congress in May, saying he proposed an AI bill to oversee the fast-growing industry, but that the technology could potentially eat away jobs. The industry is under scrutiny for its potential to spread false information.
OpenAI did not immediately respond to a request for comment.
When OpenAI first released ChatGPT in November, it quickly captivated the public imagination with its ability to answer questions, write poetry, and riff on just about any topic thrown at it. But the technology can also mix fact and fiction, or fabricate information, a phenomenon scientists call “hallucinations.”
ChatGPT is powered by what AI researchers call neural networks. This is the same technology that enables services like Google Translate to translate between French and English, and self-driving cars to identify pedestrians as they navigate the streets. Neural networks learn skills by analyzing data. For example, it can recognize cats by identifying patterns in thousands of cat photos.
Researchers at labs such as OpenAI have designed neural networks to analyze large amounts of digital text such as Wikipedia articles, books, news articles, and online chat logs. These systems, known as large-scale language models, have learned to generate text on their own, but they can repeat flawed information or combine facts in ways that produce inaccurate information. there is.
Chatbots like ChatGPT, which companies like Google and Microsoft have also deployed, are revolutionizing the way computer software is built and used. They are poised to reinvent Internet search engines like Google Search and Bing, voice digital assistants like Alexa and Siri, and email services like Gmail and Outlook.
This is a developing story. Please check the latest information.
