The Bot Battle

This article originally appeared in Intellectual Property Magazine. Reprinted with permission. Any opinions in this article are not those of Winston & Strawn or its clients. The opinions in this article are the authors’ opinions only.

The advancement of artificial intelligence (AI) is disrupting several legal frameworks, including the foundations of patent litigation. Since 1952, 35 USC Section 284 has dictated the award of damages in patent cases, “Upon finding for the [patentee] the court shall award the claimant damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made of the invention by the infringer, together with interest and costs as fixed by the court.”

But the continued growth of AI across various industries calls into question how patent damages should be calculated to compensate– without overcompensating – owners of AI-related patents.

Calculating patent damages is not an exact science and presents certain formidable challenges. Recent Federal Circuit decisions confirm that it continues to be a moving target for patent owners and accused infringers alike.

In particular, recent decisions highlight the amorphous nature of the law of apportionment. Apportionment is crucial to almost every patent damages case and applies to both lost profits and reasonable royalties. Indeed, damages calculations may vary by orders of magnitude based on different apportionment approaches applied to the same accused products.

Undoubtedly, the amorphous nature of the law of apportionment will be even more pronounced when applied to the budding world of AI. For example, AI-based medical devices are inherently complex, multi-component products for which it will be a challenge to apportion the value attributable to the patented technologies embodied in the various components of the devices. The primary purpose and function of AI-based medical devices is to perform a medical task, not to execute an AI algorithm. The devices are simply enhanced using AI algorithms, such that their primary purpose and function may be achieved more efficiently and/or more accurately.

So, under these circumstances, where an asserted patent claims an AI algorithm, should patent damages be based on the sales of the entire device or just sales of the components implementing the AI algorithm? If the latter, what if the components implementing the AI algorithm also implement non-patented functionality? The recent Federal Circuit decisions shed some light on these issues.

Continue reading the Intellectual Property Magazine article here.