{"id":1631488,"date":"2026-03-04T15:45:31","date_gmt":"2026-03-04T20:45:31","guid":{"rendered":"https:\/\/observer.com\/?p=1631488"},"modified":"2026-03-04T15:22:10","modified_gmt":"2026-03-04T20:22:10","slug":"copyright-law-age-of-ai","status":"publish","type":"post","link":"https:\/\/observer.com\/2026\/03\/copyright-law-age-of-ai\/","title":{"rendered":"Can Copyright Survive the Age of A.I.?"},"content":{"rendered":"
\"An
In classrooms and courtrooms alike, the future of copyright is being negotiated in real time. Unsplash+<\/span><\/figcaption><\/figure>\n

Think about your favorite song, movie or musical. Then consider how you enjoy, watch, listen or engage with that content. The answer? Copyright.<\/span><\/p>\n

Copyright has not disappeared in the age of A.I., but it is being stress-tested in novel ways. Take music, for example. Digital service providers (DSPs) now report tens of thousands of <\/span>A.I.-generated tracks<\/span><\/a> being uploaded daily. Some of these tracks have even risen to the top levels of the listening charts, competing alongside human-created works. Generative systems are now producing scripts, visual art and music at scale, blurring the lines between inspiration, imitation and automation. Against this backdrop, current conversations around copyright will determine whether it remains a living promise to creators or becomes an artifact that companies politely acknowledge but can quietly route around.\u00a0<\/span><\/p>\n

I have spent my career defending copyright on behalf of record labels, studios and creators. I now teach business students who are eager to both build with A.I. and protect the sanctity of human creativity. If copyright was ever at a crossroads, that moment is now. What a time to teach the future leaders of the entertainment industry about something that\u2019s changing before our eyes.<\/span><\/p>\n

A career framed by copyright\u2019s evolution\u00a0<\/b><\/h3>\n

Before joining the full-time faculty of American University\u2019s Kogod School of Business, where I am the director of the <\/span>Billboard-recognized<\/span><\/a> Business & Entertainment program, I spent decades working for companies that relied upon copyright\u2014both for protection and for innovation; copyright not only protects catalogs of creative works, but it also justifies investing in new artists, stories and songs. Without predictable ownership and licensing frameworks, there is little incentive to finance risk.\u00a0<\/span><\/p>\n

In 2013, I left corporate America precisely due to concerns that individual songwriters, filmmakers and other independent creators would be drowned out by better-funded voices when policy lines were drawn. In 2026, A.I. technologies have proven much of that worry held water<\/keyword>. The shift from CDs to MP3s, and from physical distribution to streaming platforms like Netflix and Hulu, reshaped the economics of culture. Yet even those transitions were discrete format or platform shifts. Because the impact of A.I. is broader, the stakes surrounding A.I. are higher.<\/span><\/p>\n

Unlike a new distribution model, A.I. systems can ingest vast quantities of copyrighted works in seconds, generate outputs at scale and compete directly with the very material that trained them. From this vantage point, the rise of A.I. makes stronger the underlying premise that a human author\u2019s contribution should be the anchor of the creative ecosystem.<\/span><\/p>\n

What the law is quietly signaling\u00a0<\/b><\/h3>\n

For all the chaos and uncertainty, U.S. law thus far has sent a clear message: humans are still at the center of copyright. The U.S. Copyright Office has emphasized that human authorship is required for protection; machine-only works do not qualify, a position courts have affirmed in cases like <\/span>Thaler v. Perlmutter<\/span><\/i>.\u00a0<\/span><\/p>\n

At the same time, courts and agencies are debating whether ingesting massive amounts of copyrighted material to train A.I. models equals infringement or \u201cfair use.\u201d A growing wave of lawsuits\u2014filed by authors, artists, music publishers and media companies\u2014has produced early decisions that cut in different directions. Some judges have emphasized transformative use, while others have signaled limits.\u00a0<\/span><\/p>\n

While we wait for these court opinions to help pave the legal rules of the road, some rightsholders are turning to licensing deals and settlements. These private negotiations between A.I. companies and publishers, music rights organizations and media companies hint that these training uses will be explicitly licensed rather than merely exploited in the future, the impact of which no one can really be sure of yet. These agreements may help formalize compensation structures, but they also raise questions about who has leverage to negotiate and who does not. Large catalog owners may secure deals. Independent creators often lack similar bargaining power. That imbalance represents one of the defining tensions of this moment.\u00a0<\/span><\/p>\n

The legality of A.I. model training<\/b><\/h3>\n

When students ask whether it is \u201clegal\u201d to train A.I. models on copyrighted works scraped from the internet, they are really asking a deeper question: \u201cDoes my work matter if a machine can learn from it for free?\u201d Today\u2019s business students are digital natives and emerging rightsholders. Many are also artists themselves. They are experimenting with A.I. tools to write scripts, design campaigns and prototype businesses. And importantly, I do not tell them to turn away from these technologies. They are weighing whether and how a career in the entertainment industry will be sustainable in the years ahead. I expect the presence of A.I. to continue to grow, so they need to understand it deeply to achieve impact in their respective fields of passion.<\/span><\/p>\n

Addressing the underlying trepidation around their work\u2019s value is relatively easy, but the answer to the question around legality is more nuanced and complex. The current copyright framework was not built for a world where an A.I. system can absorb millions of books or songs in an instant, then generate outputs at scale that feel close enough to the originals to devalue them. Where technology moves exponentially, the law moves incrementally.\u00a0\u00a0<\/span><\/p>\n

Rather than offering simplistic answers, I prompt students to consider three questions that are increasingly shaping responsible engagement:<\/span><\/p>\n