Published on

Executive Order by President Biden Addresses Artificial Intelligence

Insights Category
Generative AI

AI can be a powerful tool for good, but it also creates large risks, both from unintentional lack of caution and from intentional harm and deception by bad actors.

Midjourney created image of President Biden talking with an AI android.

The Executive Order is an extensive list of issues related to the use of AI and recommendations on how to safeguard that usage. The Executive Order appears somewhat more comprehensive than the EU AI Act in development in Europe, but the Executive Order is less detailed. Both the Executive Order and the EU AI Act recognize that the risks of AI must be addressed now.

At a high level, I think there are two key takeaways from the Executive Order.

First, the value of the Executive Order will depend on how it is implemented by government agencies and Congress. Even the best guidance will fail if there is not substantial additional work to turn the Executive Order into meaningful oversight.

Second, it is unclear whether the Executive Order will have much impact on intentionally bad actors. For those that want to exploit AI in socially harmful ways, governments will need more than an Executive Order or laws. We will need to have some good actors develop benevolent AI systems to find and counteract the malicious usage of AI.

Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence