Media-AI Deals Outpace Lawsuits Nearly 3-to-1 as Licensing Market Matures

The numbers tell a clear story. As of September 11, there were 299 deals between AI companies and media firms, against just 107 lawsuits, according to Axios research tracking activity since May 2020. Deals overtook litigation as the dominant mode of engagement back in June 2024, and the gap has widened steadily since.

That is not a marginal edge. It is a nearly three-to-one ratio favoring commercial agreements over courtroom fights, and it signals something important about how creative industries have chosen to respond to generative AI's appetite for their content.

Why it matters

The licensing market is developing faster than courts are establishing the underlying rules. That sequencing matters. Publishers, studios, and other rights holders are not waiting for judges to settle foundational questions about fair use, training data, and copyright infringement before cutting deals with the same companies they might otherwise sue.

This is a rational response to uncertainty, not a surrender of legal leverage. Litigation is slow, expensive, and outcome uncertain. Licensing revenue, by contrast, is immediate and negotiable. For media companies under margin pressure, a signed deal today beats a favorable ruling that might arrive years from now, if at all.

It also suggests AI firms have concluded that paying for content access is cheaper, in reputational and legal terms, than fighting every claim. Settling the commercial question through contracts sidesteps years of discovery and appeals. That calculus has clearly shaped behavior on both sides of the table since mid-2024.

The Shifting Balance of Power

Look at the trajectory. Lawsuits led the count from May 2020 through mid-2024, which tracks with the early alarm many publishers sounded as large language models scraped the open web with little regard for licensing terms. That period produced some of the highest-profile suits against AI developers, and it set the tone for how the industry initially framed the threat.

The crossover point in June 2024 marks a turning point worth dwelling on. Once deals began outpacing lawsuits, the gap did not narrow. It grew. By September 2026, cumulative deals had reached 299 against 107 lawsuits, a spread of nearly 200.

That widening gap implies more than isolated deal-making. It looks like an emerging market structure, one where licensing has become the default mechanism for resolving tension between content owners and AI developers. Lawsuits still matter, and they continue to accumulate. But they have become a secondary track rather than the primary one.

For portfolio purposes, this shift changes how investors should think about media companies' AI exposure. A publisher with an active licensing deal has a revenue line and a negotiated set of terms. A publisher relying solely on litigation has a contingent asset, one whose value depends on judicial outcomes that remain genuinely uncertain given how unsettled fair-use doctrine is for AI training data.

Key risks

Deal volume alone does not tell you about deal quality. Axios's tally counts agreements, not dollar figures, so a 299-to-107 ratio favoring deals could still mask wide variation in what publishers are actually receiving. Some licensing arrangements reportedly run into the tens of millions annually for major publishers, while others are likely smaller, exploratory, or bundled into broader partnership terms that are not fully disclosed.

There is also a legal overhang that deal-making does not eliminate. The 107 lawsuits still pending, and any new ones filed, could eventually produce rulings that reshape the economics of every existing license. If a court determines that AI training on copyrighted material without permission constitutes fair use in some circumstances, the leverage publishers currently hold in negotiations could erode quickly. Conversely, a ruling favoring publishers could make every current licensing rate look like a bargain, prompting renegotiation demands across the industry.

Concentration risk is another factor. If a small number of large AI firms account for most of the 299 deals, publishers face similar counterparty dependence to what they experienced with search and social platforms. That history, of traffic and revenue flowing through a handful of dominant intermediaries, should make rights holders cautious about how deeply they tie their business models to a similar set of players in AI.

The opportunity

For media companies, the current environment offers genuine upside. Being an early, credible licensing partner establishes commercial precedent and can command better terms than a company that waits and later gets swept into a bulk agreement. It also builds direct revenue relationships that do not depend on advertising or subscription models alone, diversifying income at a moment when both of those channels face their own pressures.

For AI companies, striking deals reduces legal risk and secures higher-quality, rights-cleared training data, which increasingly matters as model quality and provenance come under scrutiny. Paying for content also builds goodwill with an industry that has significant lobbying and public relations reach, something litigation-heavy AI firms have learned can turn costly in the court of public opinion even when they prevail in actual courts.

The Bottom Line

The market has effectively voted, and it prefers deals to lawsuits by a wide margin. That preference reflects pragmatism on both sides: media companies want revenue certainty now, and AI companies want legal certainty and clean data. Investors tracking this space should weight licensing revenue disclosures as a leading indicator of how media companies are monetizing AI exposure, while keeping an eye on the smaller but consequential set of pending lawsuits that could still rewrite the terms of every deal signed to date.