Tag: patent

Measuring FRAND Value in ZTE v. Samsung

China’s recent global FRAND determination in ZTE v. Samsung has attracted attention because it valued ZTE’s SEP portfolio at nearly twice the level of a contemporaneous English court decision. This article argues that the more important issue is not the royalty rate itself, but the methodology used to measure technological contribution. The Chongqing court relied heavily on declared SEP-family shares, portfolio metrics, territorial weighting, and other quantitative indicators that are closely associated with longstanding Chinese innovation policies encouraging patent accumulation, standards participation, and portfolio expansion. By comparing the decision to earlier Chinese FRAND jurisprudence, the English Samsung decision, USPTO research, and broader debates over Chinese patent statistics, the article explores whether these metrics accurately measure technological contribution or instead reward portfolio scale and geographic concentration. The case may signal an important shift in Chinese FRAND adjudication from disputes over the meaning of FRAND to a deeper debate over how FRAND value itself should be measured.

Post-Filing Data in Chinese Pharma Patents: Why It Took So Long — and What Finally Worked

Recently, the Supreme People’s Court of China (SPC) upheld a decision of the Beijing Intellectual Property Court reversing a China National Intellectual Property Administration (CNIPA) invalidation decision and confirming the validity of Novo Nordisk’s semaglutide compound patent. Although the final written decision has not yet been publicly released, official summaries indicate that the court accepted post-filing experimental data where “the technical effect can be derived from the original specification” (技术效果可由原说明书得出), reversing an administrative invalidation decision. Public reporting further indicates that the dispute turned on whether CNIPA would accept post-filing experimental data demonstrating semaglutide’s surprising pharmacokinetic effects in animal models, where the application as filed contained no experimental data.