The University of Tennessee Research Foundation brought a patent infringement case against Anthropic on Monday, filing in the U.S. District Court for the District of Delaware. The complaint, made public on Tuesday, claims the AI firm’s Claude Code agentic coding tool and its software architecture violate two university‑owned patents covering neuromorphic and neuroscience‑inspired computing methods.
Patents numbered 10,019,470 and 10,095,718, issued to professors Garrett Rose, James Plank, Catherine Schuman and Ahmedullah Aziz, form the core of the lawsuit. Bloomberg Law reports the foundation argues Anthropic’s products employ patented techniques for constructing neuromorphic networks, including a background execution scheduling system and a memory‑consolidation engine described in the patents.
The foundation, which manages intellectual‑property assets for the University of Tennessee’s Knoxville campus, frames the case as part of a broader pattern of what it calls a “cavalier approach” to others’ IP rights. The filing links the patent claim to Anthropic’s recent $1.5 billion copyright settlement with a class of authors and to another suit over usage limits on its Claude Max plans.
Anthropic, a fast‑growing artificial‑intelligence company, has not yet filed a formal response. Neither the company nor the foundation answered requests for comment at the time of reporting. The dispute will move forward in Delaware, a venue known for handling a large share of U.S. patent litigation.
Legal experts note that the next phase will focus on claim construction—judicial interpretation of what the patents actually cover. That process often determines whether a case proceeds to infringement analysis or is dismissed early. If the court narrows the patent scope, Anthropic may avoid liability; a broader reading could pave the way for monetary damages and an injunction that would bar further use of the contested technology.
The patents stem from research by TENNLab, a University of Tennessee group that has explored brain‑inspired computing since 2014. The lab’s early frameworks, known as NIDA and DANNA, underpin the university’s portfolio of seven issued patents and several pending applications. Schuman has emphasized the collaborative nature of the field, noting that isolated research is “basically impossible.”
Anthropic’s Claude suite, including Claude Code, is a centerpiece of the company’s strategy to expand its partner network and differentiate its offerings from rivals such as OpenAI and Google. The lawsuit underscores that intellectual‑property battles in the AI space now extend beyond copyrighted training data to the underlying algorithms and hardware designs that power large‑scale models.
While the foundation seeks unspecified monetary damages and an injunction, it has not disclosed a valuation for the patented technology nor indicated whether any licensing talks occurred before the suit was filed. The lack of a concrete damages figure leaves the financial stakes uncertain, but the potential for an injunction could disrupt Anthropic’s product roadmap.
Anthropic’s recent settlement of a massive copyright case—initially delayed by a San Francisco judge over author‑payment concerns—demonstrates the company’s willingness to resolve high‑profile disputes. However, the patent case introduces a different legal front, one that challenges the very methods the company uses to train and operate its models.
Observers will watch how Delaware judges interpret the patents and whether Anthropic can demonstrate that its systems do not embody the patented inventions. The outcome could set a precedent for how neuromorphic‑inspired techniques are protected in the rapidly evolving AI industry.
Cet article a été rédigé avec l'assistance de l'IA.
News Factory APP - actualités agentiques pour booster votre SEO et AEO.