Summary
The White House is exploring the use of cryptographic technology to authenticate official communications and combat AI-generated deepfakes. The administration has met with AI developers to discuss embedding watermarks in products, although watermarks can be manipulated. The government is developing watermarking standards through the new AI Safety Institute and the Department of Commerce. The Biden Administration has also launched the U.S. AI Safety Institute with the participation of OpenAI, Microsoft, Google, Apple, and Amazon. Some argue that using AI to detect AI deepfakes will only create an arms race, and instead suggest enabling people to prove their identity online, similar to the banking sector’s use of continuous authentication. Raising awareness and government intervention are seen as crucial in combating deepfakes.
Key Points
1. The Biden administration is exploring the use of cryptographic technology to authenticate official communications from the White House in order to combat AI-generated deepfakes. This comes after an AI-generated deepfake of President Biden aimed to deceive New Hampshire primary voters.
2. The government is developing watermarking standards through the new AI Safety Institute and the Department of Commerce to address the issue of manipulated or removed watermarks. The goal is to establish clear rules and guidance for content provenance and watermarking.
3. The Biden Administration has launched the U.S. AI Safety Institute, with participants including OpenAI, Microsoft, Google, Apple, and Amazon. This initiative was born out of President Biden’s executive order directed at the AI industry and aims to address the challenges posed by AI-generated deepfakes.