The Risk of AI-Assisted Invention And Your Company’s AI Policy
Written by Ben Esplin
Last month, I analyzed the USPTO’s new direction on AI inventorship. As I noted in that analysis, a key feature of the USPTO’s new direction is the total uncertainty regarding when or how “conception” of inventions will be determined when generative AI plays a significant role in ideation.
On the other hand, this lack of certainty should not paralyze legal and business leaders. If anything, these are the types of situations in which decision-makers can have the largest impact on an ultimate outcome. Your AI policy should not just say “disclose AI use.” It should mandate treatment of interaction history (at least in threads used to further innovation) based on educated guesses about how AI-assisted invention may be threatened through this conception-centric analysis. In a few years, a single saved prompt thread might be the only thing standing between a billion-dollar patent and a “public domain” ruling.
My recommendation is that you do not wait for the courts to draw the line. Build a policy that fits your risk profile for these types of innovations and your company’s patent policy or program in general. If you would like to discuss this further, please reach out for a consultation.
