Intro: President Biden’s executive order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence reflects the administration's comprehensive strategy for responsible innovation and its recognition of the need to strike a balance between advancing AI technologies and ensuring they are safe, secure, and beneficial for various sectors. 

Why it’s notable: 

• The executive order aims to set new standards for AI safety and security, emphasising the importance of ensuring that algorithms and models are equitable, transparent, and trustworthy. It requires that large companies share safety test results and critical information with the federal government, especially in cases where these systems could pose potential risks to national security, economic security, or public health and safety. The order also directs the National Institute of Standards and Technology to establish rigorous testing standards for AI safety, with the Department of Homeland Security applying those standards to critical infrastructure sectors.

• Additionally, the order addresses training data for large AI systems, specifically the need to alleviate algorithmic discrimination and biases.  The order  seeks to prevent discrimination in AI systems through training, technical assistance, and coordination between the Department of Justice and Federal civil rights offices. Simultaneously, it encourages the development and use of privacy-preserving techniques in AI, particularly relevant, where patient data privacy is crucial. 

Industry Implications:

• The U.S. has traditionally lagged behind other major geographies on the regulation of AI. This federal government has shown reluctance to impose regulations that might slow AI innovation or place the U.S. in a weaker position compared with countries. However, there has been a consensus that guidelines, at a minimum, are necessary. The order appears to create a framework for the use of AI, while also encouraging innovation through expansion of grants, while increasing collaboration with international partners. 

• The executive order expands upon the voluntary commitments the White House previously secured from 15 leading companies, including Microsoft, Amazon, and Alphabet,  to drive safe, secure and trustworthy development of AI. It represents the first major binding government action in the U.S. on AI technology. However, an executive order is not a permanent law, typically only remaining in effect for the duration of a sitting President’s administration. Lawmakers are still discussing how to regulate AI, with some politicians aiming to pass laws by the end of the year.  In Europe, the proposed EU regulatory framework for AI, which outlines the analysis and classification of AI systems based on risk, is currently awaiting approval.

Read the news story here.