Data licensing from publishers and creators is rising to build high-quality AI datasets, but privacy and consent concerns frequently take a back seat in these arrangements.

The push for more data access to improve AI tools raises a trade-off with user privacy, as nearly any online activity could be used for training.

Leading tech players are accelerating AI development despite ongoing legal challenges and pauses on certain models like Mythos 5.

The piece argues governments should regulate data collection practices more actively, noting that opt-out options are often hidden or hard to locate.

Overall, the article portrays the AI data race as a privacy risk and urges policy action to balance innovation with protection of individuals’ information.

At the heart of the issue is the exploding demand for data—text, video, audio, images, and user interactions—that train contemporary AI models.