Meta’s $42 billion legal bill raises questions for AI companies

New Mexico said Facebook committed 43,899,725 violations of state law and was required to pay $25 billion in civil penalties.[2] The case concerned Facebook’s conduct surrounding the Cambridge Analytica scandal and the misuse of personal information collected through the platform.[2] New Mexico also challenged the company’s representations about its data practices and its handling of misinformation and hate speech.[2] The broader issues included user privacy, third-party access to personal information and harmful content.[2] Axios reported that legal experts expected the penalty to be reduced, while Meta said it believes it will win an appeal.[1]

Axios reported that Meta agreed to a $17 billion settlement with states related to the impact of its products, including Instagram, on teen mental health.[1] Reuters described Meta’s agreement as a commitment to pay up to $18 billion over the next decade, and reported that the settlement required strict limits on how teenagers use Facebook and Instagram.[3] Reuters also reported that Meta denied wrongdoing in the settlement.[3] The evidence gives different dollar descriptions of the agreement and does not explain the difference, so readers should treat the figures as source-specific rather than assume they are identical terms.[1][3]

Axios said Section 230 protects tech platforms from lawsuits over content created by users, while AI models create their own content—potentially exposing companies to product-liability and negligence claims.[1] Potential claims involving AI could also concern harms caused by models, damage caused by people using them and copyright issues.[1] The exposure may extend beyond developers to companies that use AI.[1] Axios also warned that products may become widely adopted before courts decide whether their safeguards are adequate.[1]

POLITICO reported that states were already using consumer-protection laws against AI developers.[4] Axios said internal records of warnings, testing and product decisions could influence judges’ and juries’ findings in AI cases.[1] It also said liability risks are likely to feature in disclosures AI companies such as Anthropic and OpenAI make to shareholders as they prepare for public listings.[1]