Consumer protection in the US Executive Order on artificial intelligence.

The “Executive Order on the Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence” expressly makes consumer protection against the potentially malign effects of AI a national priority:

“The interests of Americans who increasingly use, interact with, or purchase AI and AI-enabled products in their daily lives must be protected. Use of new technologies, such as AI, does not excuse organizations from their legal obligations, and hard-won consumer protections are more important than ever in moments of technological change. The Federal Government will enforce existing consumer protection laws and principles and enact appropriate safeguards against fraud, unintended bias, discrimination, infringements on privacy, and other harms from AI. Such protections are especially important in critical fields like healthcare, financial services, education, housing, law, and transportation, where mistakes by or misuse of AI could harm patients, cost consumers or small businesses, or jeopardize safety or rights. At the same time, my Administration will promote responsible uses of AI that protect consumers, raise the quality of goods and services, lower their prices, or expand selection and availability.”

Other aspects of the EO address other aspects of consumer protection: working against invidious discrimination, enhancing privacy protections, and ensuring that competition works in consumers’ favor.

Red-teaming and sharing the results should work to the benefit of consumers.

Tyler Farrar, CISO at Exabeam, wrote to explain how the EO might be expected to serve the development of consumer protection standards. “I'm pleased to see that the Executive Order places a strong emphasis on enhancing security measures for AI systems,” Farrar said. “The requirement for sharing safety test results and conducting rigorous red team testing for sizable AI companies will help boost cyber defenses. Setting standards for content authentication and privacy-preserving techniques will also help overwhelmed security analysts as AI’s integration into security operations and other critical systems takes hold.”

Farrar concluded, “The Executive Order's commitment to protecting consumers and establishing an advanced cybersecurity program will provide security analysts with new AI-powered tools and resources to help identify and address vulnerabilities in critical software. It promises a more secure digital environment and alleviates alert burdens on analysts, helping the industry focus on addressing more pressing issues. As new AI technology matures, we are still learning how it will best integrate into workflows; this Executive Order is an important start in establishing much-needed guidelines.” 

Paul Laudanski, Director of Security Research at Onapsis, lists the challenges AI presents with respect to consumer protection.“In response to today's executive order on artificial intelligence (AI), it becomes evident that there's a substantial commitment from AI companies. However, this commitment also underscores the need for further discussions to address aspects that might not have been agreed upon initially. In assessing AI technology, it's vital to acknowledge the dual nature of its impact on consumers. While it holds tremendous promise, it also raises significant concerns, with privacy being a paramount issue. AI's capability to create highly accurate and comprehensive consumer profiles necessitates a closer look at data protection, ethical considerations, and transparency.”

He was struck by the way the EO addressed prevention of invidious discrimination. “One of the remarkable aspects of this executive order is its focus on the rental market, aimed at preventing discrimination by landlords through the use of AI algorithms. This revelation expands our understanding of AI's societal implications, highlighting its role in addressing issues we may not have previously considered.”

NIST will play a role in consumer protection. “The reference to the National Institute of Standards and Technology (NIST) spearheading penetration tests is also particularly intriguing. This initiative demonstrates how AI can be harnessed to identify vulnerabilities in code, thereby enhancing cybersecurity in critical software systems. This signifies a significant step toward safeguarding the integrity and security of our digital infrastructure.”

Laudanski concludes that the EO should enable a healthy balance of security and innovation. “This executive order underscores the importance of concrete regulation in the development of AI technology. As the world grapples with the rapid adoption of generative AI models, it's essential to strike a balance between innovation and safeguarding the public interest while maintaining transparency and clarity that doesn't shield companies or produce exceptions that may not be readily visible.”

Consumers might, in the end, benefit from more transparency.

Jeff Williams, co-founder and CTO at Contrast Security, didn’t see as much emphasis on transparency as he would have liked. “I would have liked to see much more here about AI transparency and Explainable AI. Consumers need to understand and interpret the predictions made by ML models. They have the right to know about the software and models they are trusting with the most important things in their life – finances, healthcare, government, social life, education, etc. All of this will be influenced by AI in the near term. I wish the administration had used its influence here to encourage much more transparency and explainability in the AI world.”

In some ways this is a libertarian point: an informed consumer is a protected consumer.