OpenAI CEO Sam Altman is calling for new legislation to regulate AI in a US Senate hearing. In his view, this legislation must also be constantly adapted.
In his contribution, the CEO of OpenAI indicated that the security of AI is paramount for the developer. The rollout of GPT-4 has been postponed for more than half a year to ensure that this model was reliable through testing and improvements.
More specifically in his speech, Sam Altman discussed how politicians should regulate AI, especially in the field of security. Importantly, AI developers must adhere to a clear set of security requirements. For example, conducting internal and external tests for a release and publishing these test results.
Politicians must ensure that these requirements are met in various ways. For example, through a combination of licensing and registration requirements, so that AI models have a certain standard. Companies should be given more incentives for this.
It is not entirely unimportant that Sam Altman believes that AI security legislation should be regularly adapted to developments. This requires the creation of a multi-stakeholder influence process. A large group of experts and companies have to develop and update the right safety standards on a regular basis. Also through international cooperation.
The OpenAI CEO also discussed the consequences of the technology for the labor market. He expects AI to provide a mix of more productivity for individual users and new jobs. Jobs will also change and disappear.
Read also: AI law passes important EU vote