La fonction Expert Review de Grammarly aurait utilisé un grand modèle de langage pour générer des suggestions de révision présentées sous les noms de journalistes, d’auteurs et d’universitaires sans leur consentement. Une action collective fédérale déposée par Julia Angwin a affirmé que la fonction avait détourné des identités à des fins commerciales et attribué aux personnes nommées des conseils qu’elles n’avaient jamais donnés.
Gestion pratique des risques d'entreprise et réglementations
Grammarly's automated AI writing advice signed generated articles with the names of real, prominent journalists without their permission, leading to a class-action lawsuit for false endorsements. The platform faced severe legal claims, regulatory investigations, and loss of user trust, causing immediate customer attrition. Regulatory Impact Alignment: Commercial generation of synthetic video, voice cloning, and deepfake marketing campaigns must comply with the Lanham Act (preventing false advertising and unfair competition) and EU AI Act Article 52. Any synthetic content must contain invisible cryptographic digital watermarks (e.g., C2PA standard) to prove authenticity.
Assigning synthetic content to real human authors without consent constitutes severe consumer deception and fraud. Platforms must establish explicit guidelines prohibiting fake human attribution, implement rigorous QA, and maintain transparent AI labeling. Compliance Audit Standards: For detailed verification audits, this case maps directly under Lanham Act Compliance & EU AI Act Article 52 (Synthetic Content Watermarking). Systems deploying similar AI features must maintain dynamic security logs and hold systematic compliance records.
Professional compliance incident analysis
Signing AI-generated articles with the names of real journalists is a massive compliance and consumer protection failure. It represents a direct attempt to deceive consumers. Marketing and software firms must enforce transparent AI labeling. Authenticity is not an option; it's a legal requirement.
Critical answers regarding AI compliance, auditing, and organizational risks
The writing assistant automatically generated recommendations and expert tips, signing them with the names of real journalists and authors who had never approved the text, triggering false endorsement claims.
Companies face severe prosecution under FTC guidelines for deceptive advertising and unfair commercial practices, carrying high financial penalties.
It is the policy of clearly disclosing when text, images, or insights have been fully generated or synthesized using AI, preventing customer deception.
A filmmaker reportedly used ByteDance's AI video tool Seedance 2.0 to create and post a purportedly realistic clip depicting Tom Cruise fighting Brad Pitt, which then circulated widely online. Industry groups and studios publicly alleged the tool enables unauthorized use of copyrighted material and performers' likeness, and at least one studio reportedly sent a cease-and-desist letter. ByteDance said it respects IP and would strengthen safeguards.
A Munich regional court ruled that ChatGPT reportedly reproduced protected German song lyrics and that OpenAI's models were trained on copyrighted texts, including works by musician Herbert Grönemeyer, without authorization. The court reportedly found both memorization of nine songs and lyric output to infringe exploitation rights. OpenAI disputes the ruling and may appeal. Damages were ordered, with implications for AI training on copyrighted works.
Facebook's ad delivery algorithm allegedly disproportionately showed job advertisements to one gender. Despite claims of non-discrimination, the algorithm's actions seem to perpetuate societal biases, which in turn could potentially limit opportunities for certain groups and hinder gender equity in the workplace.