AI Copyright & Generative AI Liability Litigation Explosion
Key Questions
What major copyright lawsuits are targeting AI companies?
Nearly 400 local newspapers have sued OpenAI and Microsoft, major record labels are suing Suno and Udio, authors are suing Anthropic, and investors are suing Adobe. Publishers and authors have also filed a class action against Google over unauthorized use of Google Books and Play content for Gemini training.
How does the SCOTUS Cox v Sony decision impact AI platforms?
The ruling erodes DMCA safe harbors, increasing liability risks for platforms hosting or distributing AI-generated content. This shift has accelerated litigation against generative AI developers and services.
What new legal requirements apply to AI-generated ads in New York?
New York enacted a synthetic performer disclosure law requiring labeling of AI-generated spokespeople in advertisements. It imposes compliance steps and affects vendor agreements for advertisers using such tools.
What guidance has the EDPB provided on scraping data for AI training?
The EDPB clarified that consent is not a valid legal basis for scraping personal data to train generative AI. It distinguishes between targeted and untargeted scraping as well as controller and processor roles, directly affecting ad tech firms.
What risks does the FTC settlement on fake AI services highlight?
The settlement warns against AI washing and unsubstantiated marketing claims about AI capabilities. Buried terms of service consent will not suffice, requiring rigorous substantiation for any AI-related ad tech claims.
Why are publishers suing Google over its AI training practices?
Publishers and authors allege Google used copyrighted books and Play Store content without authorization to train Gemini. Google had internally estimated licensing risks in the tens of billions of dollars.
What was the outcome of Epidemic Sound's copyright suit against Meta?
A federal judge dismissed the second suit for lack of specificity but allowed amendment. The ruling reinforces high pleading standards for mass copyright claims against platforms.
What does the NYT v OpenAI discovery sanctions motion allege?
The motion claims OpenAI hid training data and logs, potentially forcing evidence disclosure before trial. It follows the NYT dropping its secondary liability claim against OpenAI.
Nearly 400 local newspapers sue OpenAI/Microsoft; major labels sue Suno/Udio; authors sue Anthropic; investors sue Adobe. SCOTUS Cox v Sony erodes DMCA safe harbors. German court holds Google liable for AI Overview hallucinations. NYT drops secondary liability claim against OpenAI. Anthropic fair use ruling splits training vs. piracy. OpenAI's Ad Tools Terms shift liability to advertisers. New: NYT v. OpenAI discovery sanctions motion—allegations that OpenAI hid training data and logs could force evidence disclosure before trial. Google rolls out AI transparency labels for ads voluntarily, responding to FTC and EU AI Act pressure, but relies on advertiser self-reporting—a critical compliance gap. Sidley Austin piece distinguishes training vs. grounding data, offering practical risk management for ad tech AI deployment. New: New York enacts synthetic performer disclosure law—requires labeling of AI-generated spokespeople in ads, with specific compliance steps and vendor agreement implications. New: EDPB clarifies that consent is not a valid legal basis for scraping personal data to train generative AI, drawing crucial distinctions between targeted/untargeted scraping and controller/processor relationships. This directly impacts ad tech firms fine-tuning models or buying training datasets. New: FTC settlement on fake 'Active Listening' AI service warns against AI washing in ad tech—regulators will punish unsubstantiated claims about AI-powered targeting, and buried ToS consent won't cut it. This reinforces the need for rigorous substantiation of any AI marketing claims. New: Meta's AI ad tools causing brand damage—garbled images, auto-enabled features, Meta shifting blame to advertisers, highlighting operational risks of AI ad deployment and the need for transparency. New: Publishers and authors file class action against Google over Gemini training data using Google Books and Play content without authorization; Google internally estimated $10Bs-$100Bs in licensing risk. This reinforces the pattern of content owners aggressively protecting training data and threatens Google's ad tech AI partnerships. New: Epidemic Sound's second copyright suit against Meta dismissed for lack of specificity, reinforcing high pleading standards for mass copyright claims against platforms. New: AI content moderation guardrails fail at policy changes—PolicyShiftBench shows safety is policy-relative, not intrinsic; critical for EU AI Act compliance ahead of August 2 deadline for Article 50 updates.