The intricate landscape surrounding the protection of artificial intelligence (AI) inventions continues to evolve, meriting robust discussion among legal professionals. A case that neatly illustrates this topic is IBM v. Zillow, where IBM accused Zillow of infringing on seven of its artificial intelligence patents, specifically designed for assessing property values. The focus of our analysis lies in the challenges of proving patent infringement on AI algorithms and the strategy of integrating further patents aimed at secondary features of an AI system.
In the first part of this exploration, we examined these aspects and now, we’ll delve into the specifics of the claims from IBM v. Zillow, offering insights for drafting AI-related claims from a patent violation standpoint.
Claim drafting demands careful consideration of complex elements. For instance, the ‘183 patent in the IBM v. Zillow case could see some refinement. Precisely, the ‘comparing’ step of claim 1 seems to meld a training step with an inference step. This fusion might invite risks from an infringement angle. Improved claim drafting can therefore anticipate deep learning technologies and their potential infringement.
These observations highlight the significant value in diversifying a patent portfolio for AI inventions. Obtaining well-structured patent rights that circumvent future infringements form a critical defensive strategy and having a broad variety of claims can bolster it.
Learn more about the subject from the original analysis here.