The modern “boiler room” isn’t a sweaty office filled with aggressive brokers on phones. It’s an air-conditioned server room with algorithms running advertising engines.

Social media platforms are now the dominant distribution channels for pump-and-dump schemes and fraudulent securities offerings. Yet plaintiffs have repeatedly collided with the same obstacle: Section 230 of the Communications Decency Act.

For years, platforms successfully argued they were passive hosts of third-party financial promotions, resulting in summary dismissals of complaints—even for fraud.

Three decisions out of the US District Court for the Northern District of California are changing that paradigm. An especially surprising change opens the door to securities fraud Rule 10b-5 claims against social media corporations.

AI Amplification Authorship?

In Suddeth v. Meta Platforms, Inc., Chief Judge Richard Seeborg dismissed a class action brought by licensed investment advisers whose identities scammers had hijacked to promote pump-and-dump schemes in Chinese penny stocks.

Fraudsters purchased ad placements through Meta’s ad manager, uploaded fabricated endorsements using plaintiffs’ names and headshots, and funneled victims into WhatsApp groups where the manipulation continued. One promoted stock lost 90% of its market capitalization in minutes.

Plaintiffs argued that Meta’s machine-learning systems, which “maximize reach, engagement, or downstream actions,” made Meta a co-developer of the fraud. Algorithmically amplifying an illegal securities solicitation is itself content development. Seeborg rejected this theory.

Drawing on Dyroff v. Ultimate Software Group, Inc. and Doe v. Grindr Inc., he held that Meta’s targeting tools are “content neutral on their own.” Algorithmic amplification, he wrote, “is nothing more than an averment of facilitation.”

On the same day, Seeborg denied a Section 230 dismissal in Bouck v. Meta Platforms, Inc., a parallel penny-stock class action where plaintiffs alleged Meta’s generative AI tools had themselves “developed the ultimate content of the fraudulent ads,” making Meta “a genuine co-conspirator in the creation of the offending content.”

That allegation tracks the 2024 ruling in Forrest v. Meta Platforms, Inc., where Judge P. Casey Pitts allowed a near-identical theory to survive dismissal. Forrest alleged that Meta’s ad tools “mix and match” images, videos, text, and audio supplied by advertisers and use generative AI to “automatically optimize ads to versions the audience is more likely to interact with.”

Under the US Court of Appeals for the Ninth Circuit’s framework in Calise v. Meta Platforms, Inc., that active involvement in shaping the ads creates a genuine factual dispute over “material contribution” to their illegality.

The Section 230 line becomes a technical distinction: Targeting an audience is protected distribution; transforming or generating ad content isn’t. Securities law asks the same question and answers it with liability rather than immunity.

Janus Answers Back

I have argued that a platform exercising curatorial judgment over investment solicitations occupies the regulatory position of a participant in the offering, not a passive host.

The US Supreme Court’s “maker” doctrine in Janus Capital Group, Inc. v. First Derivative Traders supplies the Rule 10b-5 corollary: “The maker of a statement is the person or entity with ultimate authority over the statement, including its content and whether and how to communicate it.”

The court added that “merely hosting a document on a Web site does not indicate that the hosting entity adopts the document as its own statement or exercises control over its content.”

That sentence’s unstated implication is the argument no court has yet reached: When a platform’s generative AI exercises ultimate authority over the assembled content of an investment solicitation, the platform may be the maker of the fraudulent statement under Rule 10b-5. Primary Rule 10b-5 liability is categorical; it has no Section 230 analog.

Risk Follows AI

The Northern District of California decisions aren’t limited to Meta. Alphabet Inc., Snap Inc., TikTok Inc., and X Corp. all deploy generative AI in their advertising products.

Any platform whose tools exercise ultimate authority over assembled ad content faces the same exposure—the Section 230 material-contribution theory that survived in Forrest and Bouck, and the unresolved Rule 10b-5 maker question that Janus frames.

The Securities and Exchange Commission historically has focused its social media enforcement on individual promoters and financial influencers. Judicial recognition that platforms actively participate in generating fraudulent investment solicitations may invite scrutiny of whether those platforms are operating as unregistered broker-dealers.

For investors holding positions in ad-dependent platforms, the operative risk isn’t that Section 230 protection will disappear. For hosted content, that immunity largely holds.

The risk is that the AI tools powering their advertising products are generating content, at which point the Section 230 immunity falls away and the platform confronts a Rule 10b-5 question it has never had to answer.

The cases are Bouck v. Meta Platforms, Inc., N.D. Cal., No. 25-cv-05194-RS, opinion 3/24/26; Suddeth v. Meta Platforms, Inc., N.D. Cal., No. 25-cv-08581-RS, opinion 3/24/26; and Forrest v. Meta Platforms, Inc., N.D. Cal., No. 22-cv-03699-PCP, opinion 6/17/24.

This article does not necessarily reflect the opinion of Bloomberg Industry Group, Inc., the publisher of Bloomberg Law, Bloomberg Tax, and Bloomberg Government, or its owners.

Author Information

Seth Oranburg is a professor at the University of New Hampshire Franklin Pierce School of Law and director of the Program on Organizations, Business, and Markets at NYU Law’s Classical Liberal Institute.

Write for Us: Author Guidelines