Page 8 - Kattison Avenue Newsletter - Spring 2026 - Issue 16

this reasoning by way of example, the court observed that
if the scammers prompted Meta’s AI tool to generate “an ad
promising astronomical weekly investment returns,” and if the
AI “generated a slew of ads saying just that and new ads with
language like ‘Tired of living paycheck to paycheck? Break the
cycle and start earning steady weekly income with our proven
system,’” the scammers “did not come up with that (patently
fraudulent) language; it was all Meta.” Meaning, even if Meta’s
10
AI tool generates many versions of an advertisement that a
user may choose from, any version that includes text or images
selected entirely by the algorithm, with only “inspiration from the
scammers” (i.e., not explicitly requested by the user’s prompt), is
11
— as the plaintiffs allege — the literal creation of Meta’s AI tool.
The court’s conclusion that plaintiffs plausibly alleged that
content generated using Meta’s AI tools can be “the creation of
12
Meta” is unprecedented. Prior to the Bouck decision, no court
in the Ninth Circuit allowed a claim to proceed on the theory that
a social media platform may be liable as the “creator” of unlawful
advertising generated through its AI tool by a third party. In
fact, the court could have denied Meta immunity under Section
230 simply on the basis that plaintiffs plausibly alleged Meta
“materially contributed to the creation of the ads” generated
through its AI tool, rather than concluding that the allegations
13
supported treating the advertisements as Meta’s own creation.
The distinction between liability for material contribution to the
creation and liability as the creator is monumental: “material
contribution to the creation” sounds in aiding and abetting
liability; “creation” begets direct liability. Whereas liability for
aiding and abetting requires a plaintiff to prove that a defendant
acted with knowledge or intent to further the underlying violation
of law, depending on the cause of action, direct liability may be
14
imposed without a further showing of knowledge or intent. For
By the same logic, a social media platform should be immune
example, “intent is not a required element of a Lanham Act false
to liability under Section 230, even if the AI tool offered to
advertising claim,” but to establish that a social media platform
15
advertisers on its platform uses a GenAI model. The content
aided and abetted a violation of the Lanham Act by materially
created using GenAI depends on user input, and a GenAI tool
contributing to the creation of false advertising, a plaintiff would
functions the same way regardless of whether the content
still need to show that the platform knew its AI tool was being
prompted by the user has an unlawful purpose. Nevertheless,
7
used for false advertising. This is not easy to do and would likely
in Bouck, the court found that Section 230 did not bar claims
require showing that the platform had knowledge beyond what
against Meta Platforms, Inc. (Meta) arising from a third-party
could be derived by processing the generated content through
investment scam on Meta’s platforms where the plaintiffs
a routine review process. But to establish that a social media
16
alleged that scammers used Meta’s GenAI tools to help create
platform directly violated the Lanham Act as the “creator” of
the fraudulent advertisements.
8
false advertising, a plaintiff would only need to show that the
Accepting the plaintiffs’ allegations as true, the court concluded platform provided the AI tool that generated the ad. In effect,
that Section 230 would not shield Meta from liability if Meta’s as mentioned above, Bouck could make it easier to sue tech
AI tools “literally generat[ed], using artificial intelligence, the companies that provide GenAI tools to advertisers on their
[fraudulent] images and text in the advertisements.” Underlining platforms.
9
8 Kattison Avenue | Spring 2026