Cox and Effect: Why Volitional Conduct Is AI Copyright’s Next Battleground

Key Takeaways

• Cox decision ends most ISP contributory infringement claims
• Courts now focusing on volitional conduct for AI output liability
• Anthropic argues users, not AI, cause copying in Claude case
• Disney v. MiniMax preserves direct infringement claim based on AI design
• Volitional conduct doctrine stems from a single 1995 Netcom sentence

Pulse Analysis

The Cox Communications ruling marked a watershed moment for copyright law, effectively pulling the rug out from under the secondary liability theory that had long haunted internet service providers. By vacating the Grande Communications verdict and prompting voluntary dismissals against Verizon and Altice, the Supreme Court signaled that merely furnishing service to known infringers no longer satisfies the contributory infringement test. This rapid legal cleanup forces rights holders to look elsewhere for enforcement, and the emerging focus is on the direct infringement side of the equation—specifically, who "presses the button" when an AI system produces a protected work.

At the heart of the next battle is the doctrine of volitional conduct, a judicial construct that traces back to a single sentence in the 1995 Netcom case. Courts have traditionally required a human actor to initiate the copying, but generative AI blurs that line by autonomously selecting and remixing copyrighted material. Recent filings by Anthropic, Perplexity, and the Disney‑MiniMax dispute illustrate two opposing strategies: defendants argue that users’ prompts are the volitional act, while plaintiffs point to the developers’ training choices, system design, and marketing as evidence of corporate volition. The Ninth Circuit’s VHT v. Zillow test—examining control, selection, and instigation—provides a framework that could extend liability to AI creators even when users do not explicitly request infringing outputs.

The stakes extend far beyond individual lawsuits. If courts adopt a broad interpretation of volitional conduct, AI companies may face direct infringement claims without needing to prove inducement or tailoring, dramatically increasing litigation risk and insurance costs. Conversely, a narrow reading would leave users as the primary defendants, potentially exposing millions of consumers to liability for content they never intended to copy. Policymakers may need to intervene with clearer statutory guidance to balance innovation with copyright protection, while firms should consider embedding robust guardrails, transparent attribution mechanisms, and licensing strategies to mitigate the looming threat of volitional conduct liability.