Buchanan also discussed the potential misuse of AI, such as the creation of deepfakes, and the measures being taken to mitigate these risks, including watermarking systems and the establishment of the AI Safety Institute. He mentioned the use of the Defense Production Act to compel companies to report safety results and developmental work. Buchanan also touched on the possibility of future legislative conversations around AI, acknowledging that there are certain actions that cannot be done through executive authority.
Key takeaways:
- Ben Buchanan, the new AI advisor for the White House, discussed the challenges and progress in AI policy, including President Biden's Executive Order on AI which demands greater transparency from companies developing AI tools.
- Buchanan admitted that the White House's power to set standards for AI technology is very limited and that they do not have the legal authority to do so by executive action. However, he hinted at future regulation around AI.
- The debate around copyright and AI has intensified, with the US Copyright Office considering new rules around generative AI and the training of large language models. Buchanan stated that the White House does not have an official stance on this issue yet.
- Buchanan also discussed the potential risks and benefits of generative AI, the issue of deepfakes, and the executive order's directives for AI safety and watermarking content. He mentioned that the White House is exploring ways to cryptographically verify their own communication to ensure authenticity.