Scholars warn cheap AI reverse engineering will strain IP law

"Generative AI is putting the law of reverse engineering (RE) under severe strain," argue five scholars in a new preprint posted to arXiv in late October. The paper, by Ilan Mandel, Madiha Zahrah Choksi, Joey Castillo, Sebastian Benthall and James Grimmelmann, makes the case that existing intellectual property law "generally allows" reverse engineering but "relies on its difficulty as a filter to separate (permissible) RE from (impermissible) copying and theft." That filter, the authors say, is dissolving.

The abstract claims AI systems "can identify hardware components, extract firmware, decompile object code, and clone entire apps," making the practice "much faster, cheaper, and easier." They argue this will challenge "a wide range of laws, including trade secret, copyright, anti-circumvention, computer misuse, and contract."

On trade secrets, the authors note that protected information must be "not generally known or readily ascertainable through proper means." With AI, they write, such information is "now 'readily ascertainable' by anyone with access to a copy." On the DMCA's anti-circumvention regime, they flag what they describe as a paradox: "it is legal to use a tool to circumvent access-control technologies, but not legal to provide others with one." The piece sketches possible legal responses and their trade-offs rather than proposing a single fix.

The paper is scheduled for the ACM Symposium on Computer Science and Law in Chicago, March 23 to 25, 2027. Two of the AI researchers we track have shared the link.