US Senator Richard Blumenthal is calling for the creation of a new, independent licensing regime to determine who can develop "high-risk" artificial intelligence models. The proposal, which is co-sponsored by Republican Senator Josh Hawley, would require developers of high-risk AI models to obtain a license from a new government agency.
Blumenthal and Hawley argue that the current regulatory framework for AI is inadequate to address the risks posed by high-risk AI models. They point to the potential for these models to be used to create deepfakes, discriminate against people, or violate privacy.
The proposed licensing regime would be designed to ensure that only those with the necessary skills and knowledge are allowed to develop high-risk AI models. The new agency would also be responsible for monitoring the use of these models and enforcing the licensing requirements.
The proposal has been met with mixed reactions from the AI industry. Some companies, such as Microsoft, have expressed support for the idea of licensing high-risk AI models. Others, such as OpenAI, have argued that the proposal is too restrictive and could stifle innovation.
The debate over how to regulate AI is likely to continue in the coming months and years. Blumenthal's proposal is just one of several that have been put forward. It remains to be seen whether Congress will take action to regulate AI in the near future.
In the meantime, it is important to be aware of the potential risks posed by high-risk AI models. We should also be critical of the arguments made by those who seek to regulate AI. It is important to ensure that any regulations that are adopted are proportionate and do not stifle innovation.