ChatGPT Adds Real Cartoonists' Signatures to Fake New Yorker Art

What happened

Why it matters

• ·Enterprises deploying AI image tools face direct intellectual property and identity-falsification risk. Outputs may attribute content to named individuals who never produced it. This creates liability under right-of-publicity and defamation frameworks, even before sector-specific AI laws take effect.
• ·The incident reveals that a vendor licensing deal at the platform level does not automatically cover all outputs the model can generate. Compliance teams cannot assume a signed contract with an AI vendor settles downstream rights for every capability the tool exposes.
• ·Output integrity controls are the critical gap here. If a model can attach a real person's name or signature to fabricated content, standard content review workflows will not catch this harm. Validation processes must be extended to cover misattribution in images and other media.

Governance controls affected

What to do now

• ☐Test your organization's AI image tools by prompting them to produce content in the style of named real individuals and check whether the outputs include those individuals' names, signatures, or identifying marks without authorization.
• ☐Review existing AI vendor contracts to confirm whether the license explicitly addresses what the model is permitted to generate about or in the name of real, identifiable people, and flag any gaps to legal counsel.
• ☐Add misattribution to your AI output review checklist: reviewers should check not only for harmful content but also for outputs that falsely credit or sign a real person's name to AI-generated work.
• ☐Confirm with your AI image vendors what guardrails exist to prevent signature or identity replication, ask for documentation of how those guardrails are tested, and record the response as part of vendor due diligence.
• ☐Assess whether your acceptable-use policy for AI tools covers employee-generated prompts that request content in the style of named artists or creators, and update it if not.

What to watch next

Enforcement interest in AI-generated misattribution is growing. Courts and regulators are already active on deepfakes and identity falsification. This incident adds a documented case of a major AI platform producing content that falsely bears real creators' identities. Compliance teams should monitor whether the FTC Enforcement on AI (Section 5 of the FTC Act) or state attorneys general treat AI misattribution as a deceptive practice. They should also watch for updates to Conde Nast's or other media companies' licensing terms with AI vendors following this disclosure. The broader question of what a platform-level content license actually permits is likely to surface in pending litigation over AI training data and output rights.