Background
This decision is the latest chapter in one of the most procedurally complex patent disputes in recent memory. In October 2020, a Virginia federal judge awarded Centripetal Networks approximately $1.9 billion — at the time one of the largest patent verdicts in U.S. history — against Cisco Systems for infringing four Centripetal patents covering network-security technology. That verdict was vacated by the Federal Circuit in 2022 after it emerged that the district judge’s wife held Cisco stock, creating a disqualifying conflict of interest under 28 U.S.C. § 455.
While the district court case unraveled, parallel inter partes review (IPR) proceedings at the PTAB targeted the same patents. IPR2022-00182 was filed by Palo Alto Networks (PAN) in November 2021 and challenged U.S. Patent No. 9,917,856, which covers a packet-filtering system for detecting network threats in encrypted communications. The system correlates observable, unencrypted metadata from network packets with encrypted sessions, enabling rules-based blocking or inspection of suspicious traffic without necessarily breaking encryption end-to-end.
Cisco — time-barred from filing its own IPR petition because it had already been served with Centripetal’s complaint more than a year earlier — joined the PAN IPR as a petitioner, as did Keysight Technologies. But the Cisco connection would haunt this proceeding: one of the Administrative Patent Judges (APJs) who voted to institute the IPR held up to $15,000 in Cisco stock. Centripetal moved for recusal; the PTAB denied the motion and sanctioned Centripetal for raising it. The Federal Circuit reviewed the recusal issue and, in October 2025, ordered a new panel of APJs to hear the case — lending Centripetal a rare second opportunity to argue invalidity before unbiased adjudicators.
The Court’s Holding
The new PTAB panel — appointed in March 2026 after the USPTO declined to terminate the IPR but agreed to replace the conflicted judges — issued its Final Written Decision on Remand on June 12, 2026, again finding the challenged claims of U.S. Patent No. 9,917,856 unpatentable as obvious.
The prior PTAB finding had been vacated on remand partly because the Federal Circuit (in a December 2024 decision in a related proceeding, No. 2023-1636) found that the Board had analyzed prior art references in isolation rather than as a whole, and had failed to make adequate findings on the motivation to combine those references. The new panel’s task on remand was to supply those missing analytical steps — not to reconsider whether the prior art itself disclosed the claimed elements. That the new panel again reached unpatentability suggests the substantive invalidity case against the ‘856 patent is strong on the merits, independent of the original panel’s composition.
Key Takeaways
- Patent owners facing an APJ conflict of interest should know that winning a recusal/new-panel order may not change the ultimate outcome if the underlying prior art case is strong — as Centripetal learned here.
- APJ financial conflicts remain a live issue at the PTAB. The original panel’s sanction of Centripetal for raising the recusal motion, later reversed by the Federal Circuit, illustrates how high the stakes can be when APJ independence is compromised.
- The invalidation of the ‘856 patent removes one of the four patents on which the original $1.9 billion verdict was based. Three of the other four patents were found non-infringed in the district court remand proceedings; the ‘856 patent’s PTAB invalidation now forecloses its use even if Centripetal attempted to reassert it.
- This is a Final Written Decision — Centripetal may appeal to the Federal Circuit, extending the litigation further. But two rounds of PTAB invalidity findings make an appellate reversal on the merits unlikely.
Why It Matters
The Centripetal v. Cisco saga has tested the limits of both the patent litigation system (the judicial bias controversy that voided a $1.9 billion verdict) and the PTAB system (the APJ bias controversy that required a new panel). The June 12 decision represents a form of institutional stability: despite years of procedural turbulence, two independent panels of expert administrative judges — the original and the court-ordered replacement — reached the same conclusion about the ‘856 patent’s validity. That convergence suggests the prior art teaching the claimed network-threat-detection techniques was clear and substantial.
For cybersecurity patent holders and practitioners, this outcome is a reminder that encrypted-traffic network security patents, while commercially important, face serious prior art risk from the rich literature of enterprise firewalls, intrusion detection systems, and encrypted traffic analysis predating 2016. Companies asserting such patents in IPR proceedings should expect scrutiny of the motivation-to-combine requirement, which remains one of the PTAB’s most exacting standards — and the one that sent this case back for a second round.