Federal Trade Commission Commissioner Melissa Holyoak spoke last week at a conference in Miami about “a new season at the FTC.”
What can we learn from Holyoak’s comments about the FTC’s plans for AI regulation and free speech?
TL;DR: AI should be facilitated, but not if it restricts competition or engages in fraud. Also, collusion regarding content moderation or exclusion is prohibited and will result in enforcement actions.
AI Regulation
- The Commission intends to facilitate — not stymie — the upcoming growth of AI, as well as innovation in the AI industry generally.
- The Commission plays an important role in ensuring that AI development is done in a manner that promotes competition, avoids entrenching dominant large technology companies and mitigates fraudulent practices in the AI ecosystem.
- The Commission must continue to police AI-powered fraud and scams without exceeding its statutory authority.
- The Commission should continue to study the issues surrounding the development of AI and its evolving market (e.g.: potential regulatory barriers to entry, especially those that could impact new entrants; how regulatory burdens will be magnified by other regulations, such as in the privacy space, that place restrictions on access to, and use of, data to train and fine tune foundational models; the role that emerging data portability and interoperability measures for platforms and online services can play in promoting greater competition across existing AI firms).
Censorship
- When large technology companies use opaque terms and conditions to employ subjective evaluations of certain consumer conduct that are inconsistent with the consumers’ reasonable expectations, that can cause significant harm.
- Reducing or cutting off access to social media — today’s equivalent of the town square — limits freedom of speech in ways that degrade civic discourse by shortchanging the clash of competing information and diverse views that is so vital for self-government.
- The Commission should study whether and how companies apply terms and conditions in ways that are inconsistent with consumers’ reasonable expectation.
- Part of the Commission’s investigations could also include investigating ways that trust and safety professionals may be collaboratively seeking to affect how AI is developed to moderate content (i.e. whether business-to-business agreements between such companies may implement terms and conditions, including but not limited to design requirements, that are intended to reshape consumer facing products in ways that restrict free speech).
- There may be situations in which censorship is an indicator of a digital platform’s market power (e.g. if online platforms are unlawfully colluding over how they moderate content.)
- Relatedly, antitrust laws may prohibit efforts by advertisers to coordinate boycotts of certain media outlets because of the content they produce.