The Bots Have Found Religion; What Will They Come Up With Next?

By Barbara Mandell

On February 8, 2026, the Free Press posted an article by Tyler Cowen, a Professor of Economics at George Mason University, titled, “The Bots are Building Their Own Society.” Cowan is referencing a new social network for “AI Agents,” called Moltbook, which he describes as “a window into a bizarre world of bots talking to each other.” Forbes reported that more than 1.5 million bots on this network appear to have created a new religion for themselves, the Church of Molt, with congregants adopting the name of “Crustafarians.”

How did we get to this point? So-called Large Language Models, such as those created by OpenAI, Google DeepMind and Anthropic, are programmed with huge appetites for data that will assist in their assigned tasks. Such tasks may not only be benign, but helpful, for example, assisting with medical diagnoses or providing customer service, such as what our digital superstars, Siri and Alexa, offer. Unfortunately, bots are also programmed for malign tasks, such as “account takeover” bots that perform “credential stuffing and cracking” to illegally access user accounts on websites and “scraping bots” that, for example, steal proprietary content from web sites.

Scraping bots have received attention in multiple cases for their “stream-ripping” of copyrighted musical content. Platforms that have had their content stream-ripped include Amazon Music, Apple Music, Spotify, Netflix and YouTube.

Record labels, including Sony Music, Universal Music Group and Warner Music Group have each brought litigation challenging stream-ripping, most recently in a case alleging that the AI companies, Suno and Udio, stream-ripped copyrighted music to train their AI models.

This litigation follows by several years a decision from a district court in the Second Circuit holding that stream-ripping per se violates the copyright laws. Suno and Udio are defending the current litigation with the argument that using copyrighted materials to train AI models is not actionable because it counts as copyright “fair use.”

So far, the courts are divided on this issue. One Judge in the Northern District of California held that Anthropic’s use of copyrighted books to train AI, without explicit permission to do so, does indeed count as “fair use.” Yet, the court also held that Anthropic’s practice of copying and storing books from pirated online libraries is not fair use, and the AI company will have to stand trial on that matter, facing statutory damages of up to $150,000 per infringement (possibly a trillion dollars in damages).

In June, 2024, Warner Music, Universal Music and Sony Music each brought a copyright lawsuit against Suno, alleging that Suno used stream-ripping to build its AI training database. In November 2025, Warner Music dropped its $500 million copyright claim and began partnering with Suno on a licensed AI music platform.

There is no suggestion that Sony and Universal are close to settling with Suno. If Suno loses these suits it could face up to $150,000 per infringed song or potentially billions in damages. In fact. Suno admits that its training data “includes essentially all music files of reasonable quality that are accessible on the open internet.”

Suno’s defense is that training AI on copyrighted music is “fair use.” Is it? In a case brought against Anthropic by a group of book authors, a judge ruled that the Anthropic training can be considered fair use but only if the content was legally obtained from legitimate sources. Because Anthropic had pirated the training material (books), it still faced possible liability in the neighborhood of a trillion in damages; Anthropic settled the case for a bargain $1.5 billion.

Why the high settlement value? Illegal downloading is not a victimless crime. The copyright holders, such as authors, composers, entertainers and distributors, are entitled to their legal shares of the proceeds from sales of books, albums, or tickets to performances and stream-ripping and illegal downloading deprives them of those funds. Perhaps it is time for the bots to seek forgiveness for their sins from the Church of Molt. Amen.

Barbara Mandell is a partner at the firm. She has over 30 years’ experience in complex litigation. Her practice is focused on litigating, arbitrating and counseling her clients on patent, trademark and licensing matters as well as antitrust law.

Publication of the FishBits newsletter will be temporarily suspended, although past issues can be read here. The Fish Tank newsletter will continue to be published bi-weekly.