AI & IP
    Back to Insights

    Who Owns What Your AI Was Trained On

    AICopyrightRisk Management

    Anthony Clemenza, Founding Partner

    Admitted in New York, 2009 · nearly twenty years in practice

    · Updated · 8 min read

    Share

    AI tools now sit inside core business operations, which raises real copyright questions about the data those tools were trained on.

    Where Things Stand

    Generative AI has been deployed faster than regulators can write rules, leaving companies in uncertain legal territory. The key questions include:

    • Training data provenance: Understanding where your AI vendor’s training data originated
    • Output ownership: Clarifying who owns AI-generated content and under what conditions
    • Liability allocation: Ensuring contracts properly allocate risk for IP infringement claims

    Practical Steps for Risk Mitigation

    1. Audit your current AI tool usage across the organization
    2. Review vendor agreements for data handling and IP provisions
    3. Implement internal policies governing acceptable AI use cases
    4. Document your good-faith compliance efforts

    Looking Ahead

    The law here will keep changing. Companies that put governance frameworks in place now will adapt more easily as the rules settle.

    Found this useful? Share it.

    Share

    By email

    The Private Brief, in your inbox

    One piece at a time, whole, on the day it is published. Not a digest, not a roundup, and not a monthly summary of things you have already seen.

    Your address is used for this and nothing else, and every issue carries an unsubscribe link that works on the first click.

    Need Guidance?

    Let’s discuss your situation.

    Request an Introduction