The 1st Civil Chamber of the TJ/DF upheld a ruling denying compensation to a doctor whose image and voice were altered using artificial intelligence in a sponsored Instagram video falsely recommending a dietary supplement.
The lawsuit sought to hold Facebook and the alleged manufacturer of the product featured in the altered video accountable.
New STF rules don’t apply retroactively
The court found that while the video was illegal, the new rules established by the STF for platform liability regarding paid ads and promotions don’t apply to this case. The incident occurred before the STF established its rulings on Themes 533 and 987, which were explicitly made non-retroactive.
The doctor, an endocrinologist, gave an interview in June 2023 discussing weight-loss medications and warning against their use without medical supervision.
In March 2024, she discovered a manipulated version of the interview posted as a sponsored ad on Instagram. Deepfake technology altered her words to make it seem like she endorsed a dietary supplement whose commercialization had been suspended by Anvisa.
She sued Facebook and the company she pointed to as the manufacturer of the product, demanding the video’s removal, data to identify the poster, and R$20,000 in damages.
Lower court partially granted requests
The lower court partially granted her requests, ordering Facebook to remove the video and provide identifying data but denying compensation from the companies.
On appeal, the doctor argued that Facebook should be liable because the post was paid for and because multiple reports were made to the platform before the video was removed. She claimed the paid promotion amplified the fraud’s reach.
Regarding the other company, she asserted that it was responsible for the supplement’s manufacturing and would have benefited economically from the disclosure.
Read Also: IASP launches course on oil and gas dispute resolution
TJ/DF clears Facebook of liability
Reporting judge Teófilo Caetano emphasized that the STF, in Themes 533 and 987, partially struck down Article 19 of the Marco Civil da Internet and set new parameters for provider liability.
These include presumed liability for illicit content in paid ads or promotions, even without prior notice. However, the STF modulated the effects of the decision to be applied only prospectively.
Since the video was posted in March 2024, the judge ruled that the case remains subject to the rules in force at the time of the facts. Thus, he applied the regime of Article 19 of the Marco Civil da Internet then in effect, under which the provider could only be held liable if, after a specific court order to remove the content, it remained inactive.
In this case, there was no prior court order when the lawsuit was filed. Although Facebook exceeded the initial 24-hour deadline set in the urgent relief, it removed the content after a new decision that imposed a daily fine for non-compliance. The judge found no grounds to attribute illicit conduct or causal link to the platform regarding the alleged damages.
The judge also considered unproven the reports that the author claimed to have made previously to the platform.
Regarding the company pointed to as the manufacturer of the supplement, he concluded that there was no evidence of its participation in the editing, publication, or promotion of the video. Nor was it proven that the company obtained increased profit from the post or had a causal link with the illicit act.
With these grounds, he voted to deny the appeal and maintain the sentence.
Dissenting opinion
Rômulo de Araújo Mendes partially dissented. While acknowledging the prospective application of the new STF theses, he argued that Facebook’s liability was independent of this understanding, based on the specific circumstances of the case.
The chamber thus upheld the original ruling by a majority vote.
