%20(1).png)
OpenAI, the firm behind ChatGPT, wants the US to explore legislating licences for businesses developing sophisticated artificial intelligence like the sort that powers its chatbot, according to its CEO, who prepares to testify before Congress on Tuesday.
According to his written statement, CEO Sam Altman is slated to push for licencing or registration restrictions for AI with limited skills in his first appearance before a legislative panel. In this approach, the US may hold firms accountable for safety regulations, such as testing systems before to release and reporting the findings.
"Regulation of AI is essential," Altman said in prepared remarks seen by Reuters.
For months, corporations of all sizes have rushed to bring increasingly dexterous AI to market, dumping mountains of data and billions of dollars at the task. Some detractors believe that the technology would increase societal problems such as bias and disinformation, while others predict that AI could terminate mankind.
The White House has invited leading technology CEOs, including Altman, to discuss artificial intelligence. US politicians are also looking for ways to expand the technology's advantages and national security while restricting its misuse. The likelihood of agreement is remote.
According to Reuters, an OpenAI employee recently advocated the formation of a US licencing agency for AI, which may be dubbed the Office for AI Safety and Infrastructure Security, or OASIS.
In his written evidence, Altman did not mention OASIS, but he did argue for "a governance regime flexible enough to adapt to new technical developments" and "regularly update the appropriate safety standards."
Licences, according to technology experts, risk pushing out smaller businesses or losing relevance if AI grows too rapidly, but they would help the US focus monitoring and defend against exploitation.
Microsoft is a supporter of OpenAI. Altman also advocates for worldwide AI collaboration and incentives for safety compliance.
According to the chairs of the Judiciary Committee's subcommittee on Privacy, Technology, and the Law, Tuesday's meeting is a significant step towards US supervision.
"AI is no longer science fiction or fantasy." It is real, and the implications for both good and evil are very evident and present," said Senator Richard Blumenthal, the chair of the panel. He believes that AI must not lead to an explosion of disinformation and identity theft, and that the sector must be transparent and accountable.
Christina Montgomery, International Business Machines Corp's chief privacy and trust officer, will also testify on Tuesday. She is anticipated to encourage Congress to focus regulation on areas that have the greatest potential for social harm.