Bill Gates calls on US to introduce mandatory oversight of AI companies

Bill Gates calls on US to introduce mandatory oversight of AI companies
Photo: Bill Gates / European Union, 2018

Microsoft co-founder Bill Gates has called on US authorities to introduce mandatory rules for developers of the most powerful artificial intelligence systems. In his opinion, voluntary measures by tech companies themselves are no longer sufficient to control the risks arising from the rapid development of AI.

Gates said this in an interview with NBC News. He believes that legislators and law enforcement agencies should be directly involved in defining safety and monitoring requirements. In his view, compliance with such rules will create additional costs for the industry, but should not significantly slow down technological development.

Gates drew attention not only to the long-term scenario of losing control over more advanced systems. He called the use of existing AI tools by people or organizations intent on causing large-scale harm a much closer risk.

The discussion about government oversight has intensified as leading American AI companies themselves increasingly talk about the need for external rules and a possible slowdown in model development if dangerous capabilities emerge.

Anthropic believes that requirements for developers should be tightened as models' capabilities increase. For the most powerful systems, the company considers external testing, incident reporting, and deeper government oversight necessary. Anthropic is also working on metrics that should allow independent experts and authorities to track the pace of development of advanced models.

OpenAI has recently taken a similar position. The company has advocated for mandatory national safety requirements for AI in the US and international standards that would define the procedure for evaluating model capabilities, risks, and the sufficiency of safety mechanisms. OpenAI allows that under certain conditions such rules may provide for slowing down or stopping further scaling of model capabilities.

Against this backdrop, there is no complete unity among tech companies on the form of oversight. Some industry leaders support common mandatory standards, while others believe labs should decide on their own when to pause development for additional safety checks.

Gates' position is that such decisions should not remain solely at the discretion of companies that are simultaneously competing with each other to create more powerful models. He supports a system in which basic monitoring and protection requirements are set at the state level.

Sources: NBC News, Anthropic, OpenAI

News

Loading data...
Could not load the data. Try again later or contact support!
Loading data...
All News