Complaint filed with the Federal Trade Commission calls for stricter government oversight of AI text generation tools.
The Center for AI and Digital Policy (CAIDP), an AI ethics group, has filed a complaint with the Federal Trade Commission (FTC) accusing OpenAI of violating consumer protection rules in the rollout of its AI text generation tools. The complaint follows the publication of a high-profile open letter calling for a pause on large generative AI experiments, signed by CAIDP President Marc Rotenberg and a number of AI researchers and OpenAI co-founder Elon Musk.
The complaint highlights potential threats from OpenAI’s recently announced ChatGPT-4 generative text model, including the production of malicious code and highly tailored propaganda, as well as biased training data resulting in baked-in stereotypes or unfair race and gender preferences in hiring. The complaint highlights notable privacy lapses in OpenAI’s product interface, one of which is a recent glitch that potentially revealed ChatGPT histories and payment information to other users.
Consumer Harm and Deceptive Trade Practices
CAIDP contends that although OpenAI has acknowledged the potential risks associated with AI text generation, GPT-4 goes beyond what should be considered acceptable in terms of consumer harm, necessitating regulatory action. The complaint seeks to hold OpenAI accountable for breaching Section 5 of the FTC Act, which forbids trade practices that are both unfair and misleading. It characterizes AI hallucinations, where generative models fabricate facts with confidence, as a form of deception. The complaint warns that ChatGPT could encourage dishonest commercial statements and advertising, which may fall under the FTC’s jurisdiction.
Calls for Government Oversight and Independent Assessments
The complaint filed by CAIDP requests that the FTC puts a stop to any additional commercial deployment of GPT models, mandates independent evaluations of the models prior to future rollouts, and establishes definitive regulations for generative AI systems. Additionally, it urges the creation of a reporting mechanism, akin to the one that permits consumers to report instances of fraud that is openly available to the public.
FTC’s Efforts to Regulate AI
The FTC has demonstrated a keen interest in overseeing AI tools and has cautioned that prejudiced AI systems may result in enforcement measures. During a recent event held in conjunction with the Department of Justice, FTC Chair Lina Khan stated that the agency would be on the lookout for indications of prominent established tech firms trying to stifle competition. Should the FTC launch an inquiry into OpenAI – a significant participant in the generative AI race – it would represent a significant intensification in the agency’s endeavors to regulate AI.
Conclusion
The complaint filed by CAIDP against OpenAI highlights the potential risks and harms associated with the deployment of AI text generation tools. It calls for stricter government oversight and independent assessments of generative AI models before commercial deployment. The complaint also seeks to hold OpenAI liable for violating consumer protection rules and deceptive trade practices. The FTC’s interest in regulating AI tools and enforcement actions against biased AI systems further underscores the importance of ethical AI development and deployment.
