On October 25, 2024, the U.S. Copyright Office issued its final rule in the ninth triennial exemption process under the Digital Millennium Copyright Act (DMCA). This process evaluates applications to allow circumvention of technological protection measures (TPMs) on copyrighted works every three years.
Among the notable decisions, the Office denied a proposal by Jonathan Weiss seeking an exemption to circumvent TPMs on generative artificial intelligence platforms. The goal was to enable security investigations aimed at identifying and addressing inherent biases in these AI systems. While the Office recognized the importance of AI reliability research, it concluded that such an exemption would not solve the problems identified, as the restrictions were not due to TPMs, but to the policies and terms of service set by the software vendors.
In addition, the Office approved or expanded exemptions in four categories, including: class 3(a) and 3(b): expanded exemption for audiovisual and literary works for text and data mining purposes for scholarly research and teaching. This allows researchers affiliated with other non-profit institutions to access Corpora for independent research and modifies provisions related to security measures and viewing of content within a corpus. These decisions reflect the balance the Office seeks to strike between protecting copyright and facilitating research and fair use of protected works.
