Executive Order on AI: What Steps In-House Counsel Should Take Now – New Technology

Executive Order on AI: What Steps In-House Counsel Should Take Now – New Technology

To print this article, all you need to do is register or log in to Mondaq.com.

On October 30, 2023, the White House published the “Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence” (the “Order”). The scale, scope, and goals of the Order are sweeping and far-reaching, tapping into the Department of Homeland Security, the Department of Energy, the Department of Defense, the Department of State, the Department of Justice, the Department of Transportation, and the Department of Education, among others, to evaluate and develop a framework for controlling and regulating the development, use, and use of artificial intelligence across As many industries and fields as possible.

The order includes a number of sections that address eight guiding principles and priorities, including those related to privacy, civil rights, safety, and security, among other topics.

These departments set deadlines and goals for various departments to create rules, regulations, programs, and even information collection requirements for AI management. Targets for creating rules, regulations, and information collection requirements include businesses, government agencies, health care systems, the justice system, and infrastructure. The goal is to be as comprehensive as possible, and to be prepared for any developments and applications of artificial intelligence.

Although the shortest time frame is listed at 30 days from the order, the majority of guidance provides 180 and 270 days from the order as initial deadlines. It is expected that broad, comprehensive requirements will be on the horizon and may be imposed on the use and development of AI in the public and private sectors.

What steps should be taken?

For clients managing the use and development of AI, it is essential that you continue to monitor the outcome of the matter and take steps to comply.

Some proactive steps customers can take in their AI implementation include carefully managing and documenting model creation, designing models to provide constraints on use, and carefully managing training data, including knowledge of the attribution of training data. Some developers may consider collecting their own data for use in training and model development to avoid potential pitfalls.

Presumably, the most stringent requirements will be imposed on any “dual-use base model.” This is defined as a model that is trained on large-scale data, includes self-supervision, includes billions of parameters, is broadly applicable, and exhibits a serious risk to security, public health or safety, or any combination thereof. Extreme caution should be exercised if your AI development falls into this category.

Intellectual property advisors will eagerly await the USPTO’s proposed guidance on the invention and use of artificial intelligence (120 days from the order), guidance on the intersection of artificial intelligence and intellectual property (270 days), and an upcoming study on the copyright issues raised by artificial intelligence (270 days day (days or later).

The content of this article is intended to provide a general guide to the topic. It is recommended to seek advice from specialists in your circumstances.

Popular articles about: Technology from the United States

Coding with AI in an open source world
Goodwin Procter LLP
Companies should consider taking three steps to remain compliant with open source requirements when using AI to create code for their products or other purposes.

You may also like...

Leave a Reply

Your email address will not be published. Required fields are marked *