US President Joe Biden has a new one supreme command drawn. The text provides guidelines for companies engaged in the development of AI.

The White House briefing room has a new one supreme command published that American President Biden has signed. The text provides guidelines for AI companies to develop the technology safely. The introduction emphasizes that the US not only wants to reduce risks, but also exploit the benefits of AI.

Security

Developers of the strongest AI systems are being asked to share their security test results and other important information with the US government. Companies that produce foundation models that could pose a threat to national security, the economy or health must therefore provide training.

It National Institute of Standards and Technology will set strict standards in cooperation with state security. New tools and tests are also being developed for safer and more reliable AI. In addition, concrete new standards must be introduced for companies developing AI for (dangerous) biological materials.

To better protect American citizens from fraud, standards and guidelines are being implemented to better detect AI-generated content and confirm authentic content. There will also be an expanded cybersecurity program.

The goal is also a memorandum with which American military and intelligence agencies can work together. This is prepared by the White House and Homeland Security.

privacy

Federal support for the faster development of privacy-protecting technologies is a priority. There will also be one Research coordination network Strengthen technologies such as cryptography.

The way different authorities collect public data is being evaluated. Data protection guidelines for federal authorities are also being tightened. There will also be evaluations of privacy protection options for these authorities. This includes AI systems.

For the people

There are also new guidelines on civil rights and for groups such as students or employees:

• Guide for organizations to prevent AI algorithms from being used to discriminate
• Better combat discrimination with AI algorithms
• Fairly monitor AI to ensure fair usage
• Accelerating responsible AI in healthcare
• Potential of educational forms of AI
• Guidelines for maximizing the benefits of AI for workers
• Monitoring the impact of AI on the labor market

The text also emphasizes that the US wants to promote innovation and competition in AI, but wants to cooperate globally for globally safe AI. The White House wants to apply this attitude nationwide. The framework that the EU recently proposed could potentially be such cooperation. Given its relations with China, this country will undoubtedly not be part of these intentions.