The U.S. Copyright Office has released a new report indicating that generative artificial intelligence companies may be infringing copyright laws by using creators’ works without permission for training their models. The report suggests that while some uses of copyrighted material could be considered fair use, many AI companies risk crossing legal boundaries, especially when their outputs compete with the original works. The report highlights that the distinction between research and commercial use is crucial. For instance, utilizing copyrighted works for research purposes may not substitute for the original content, while commercial applications that generate similar expressive works do not fall under the protection of fair use. The U.S. Copyright Office emphasized that the transformative nature of AI training relies heavily on the context, source, and purpose of the materials used. Following the report’s release, there was significant political fallout, including the dismissal of the office’s director, Shira Perlmutter, just one day later, raising concerns about the implications for copyright enforcement in the age of AI.
The Federal Trade Commission has delayed the enforcement of its “click-to-cancel” rule, which mandates that consumers should be able to cancel subscriptions as easily as they can sign up for them. Originally set to take effect on May 14, the new enforcement date is now July 14, allowing companies more time to comply with the regulations. The click-to-cancel rule, also known as the Negative Option Rule, aims to prevent companies from creating complicated cancellation processes that differ from the signup procedure. The FTC has stated that the delay was necessary to assess the compliance burdens on businesses. The commission voted unanimously for the postponement, although two commissioners were absent due to dismissal by former President Donald Trump. This rule is intended to enhance consumer rights and simplify the cancellation process for various subscriptions across industries.
The U.S. Food and Drug Administration has appointed its first Chief Artificial Intelligence Officer, Jeremy Walsh, who previously worked as Chief Technologist at Booz Allen Hamilton. In this newly created position, Walsh will oversee artificial intelligence and information technology initiatives at the agency, aiming to enhance efficiency in operations, particularly in light of recent workforce cuts that included key technology personnel. Walsh’s background includes extensive experience with various government agencies, and he has been involved in projects related to cloud networks and data analytics. The appointment comes as the FDA seeks to modernize its approach to technology and improve its regulatory processes, particularly in the wake of challenges posed by the COVID-19 pandemic and ongoing scrutiny of government contract expenditures. Walsh’s arrival at the FDA may mark a shift towards integrating advanced technology solutions to streamline operations and improve regulatory efficiency, particularly in the management of databases across different agency centers.
Why do we care?
The U.S. Copyright Office’s report marks a pivotal moment for the generative AI industry—and the fact that its director was dismissed a day later signals that this isn’t just a legal issue, it’s a political one. For IT service providers, particularly those offering AI solutions or integrating LLMs into client workflows, this creates a landscape of increased legal ambiguity and compliance risk. you now have to ask: what is the provenance of the training data? Clients will eventually ask, regulators will definitely ask, and legal exposure is now a real and growing concern.
The FTC’s delay in enforcing the “click-to-cancel” rule until July 14 buys some time—but don’t mistake this as optional. Many MSPs and SaaS providers still operate complex or intentionally vague cancellation processes, especially around bundled services, auto-renewals, or long-term contracts. This rule is an enforcement stick that puts customer experience and ethical billing front and center.
One could argue that none of these stories have immediate teeth. The Copyright Office report doesn’t carry enforcement weight—yet. The FTC rule is delayed. The FDA appointment is one person. But brushing them off misses the point: these are indicators of institutional alignment around AI risk, consumer fairness, and regulatory modernization.
Forward-looking IT service providers should act now—not react later.