Schatz demands answers from OpenAI in controversial ChatGPT ‘spring update’

US Sen. Brian Schatz (D-Hawai‘i) is pressing OpenAI — the company behind ChatGPT — to respond to whistle-blower claims that the company “rushed” through safety tests and barred employees from warning the public about alleged risks of its latest artificial intelligence (AI) model, GPT-4 Omni, which released on May 13.

On Monday, Schatz led five lawmakers in a letter to OpenAI CEO Sam Altman asking him to outline how they plan to meet “public commitments” to ensure its AI does not cause harm, such as teaching users to build bioweapons or helping hackers develop new kinds of cyberattacks, per The Washington Post.

According to its website, OpenAI’s latest model, GPT-4o (“o” for “omni), gives free users access to features such as GPT-4 level intelligence, responses from both the model and the web, data analysis and chart building, image recognition, file uploading and ‘memory’ to provide a more helpful experience. The company said it “squeezed” through safety tests in a single week.

On Tuesday, Schatz met with White House Office of Science and Technology Policy Director Arati Prabhakar to discuss the potential benefits and emerging safety concerns from AI. Later in the day, Schatz participated in a discussion at the Center for AI Safety DC launch reception to discuss his work on AI legislation.

“AI presents us with exciting new possibilities, but it also carries immense risk,” said Schatz. “As the AI industry rapidly grows, it’s critical that we work to ensure the technology is safe for people to use and that companies are holding up their end of the bargain by following through on their safety commitments.”

Schatz, who sits on the Senate Commerce, Science, and Transportation Committee, has been a leader on AI legislation. In 2018, he first introduced the AI in Government Act, bipartisan legislation that would seek to improve the federal government’s use of AI. It was signed into law in 2020, making it one of the first AI bills to be enacted. Since then, Schatz had led legislation to provide more transparency on AI-generated content and protect people from scams through labeling and help improve extreme weather forecasting using AI technology.