Meta Seeks Dismissal of Lawsuit Alleging Pornographic Video Use in AI Training

• ​Meta denies using adult films for AI training, calling the lawsuit baseless and speculative.
• Strike 3 claims Meta used 2,400 pirated videos; Meta disputes evidence and download patterns.
• Meta argues limited IP activity was personal use, not linked to corporate AI model development.

Meta Platforms has asked a U.S. court to throw out a lawsuit accusing it of downloading and using thousands of adult films to train its artificial intelligence models. The filing, submitted Monday to the U.S. District Court for the Northern District of California, argues that the claims by Strike 3 Holdings lack factual support. Meta stated that no evidence exists showing its AI systems were trained on or contain any of the copyrighted material in question.

The company called the allegations “nonsensical and unsupported,” maintaining that the complaint is built on speculation rather than proof. The motion, first reported by Ars Technica, marks Meta’s formal denial of what it describes as “bogus” claims.

Meta Background and Allegations

Strike 3 Holdings, a Miami-based producer of adult content distributed under brands such as Vixen, Blacked, and Tushy, filed the lawsuit in July. The complaint alleged that Meta used both corporate and disguised IP addresses to torrent nearly 2,400 of its films since 2018 as part of efforts to build multimodal AI systems.

In response, Meta’s filing challenged the scale and pattern of the alleged activity. It stated that only 157 of Strike 3’s films were reportedly downloaded using Meta’s corporate IP addresses over seven years, averaging about 22 downloads per year from 47 separate addresses. Meta’s attorney, Angela L. Dunning, described this as “meager, uncoordinated activity” likely carried out by individuals for personal reasons, not as part of an AI training operation.

Disputed Evidence and Network Scope

Meta also disputed Strike 3’s claim that over 2,500 “hidden” third-party IP addresses were linked to the company. The filing noted that many of those addresses were unverified or registered to unrelated entities, including a Hawaiian nonprofit with no connection to Meta.

The company added that there was no indication it knew about or benefited from any alleged downloads, emphasizing that constant monitoring of its global network is neither feasible nor legally required.