Israeli Patent Practitioners Warn Careless AI Use Can Erode Patent Rights After DABUS Ruling

Published October 27, 2026

Score

Why it matters

A practitioner essay published on NatLawReview on 27 October 2026 by Kfir Luzzatto and Lilach Luzzatto Shukrun of Luzzatto Group argues that AI is a powerful but indiscreet assistant. It is useful for mechanical and administrative tasks but can cause damage when used during inventive activity or strategic claim drafting. The authors map the risks across confidentiality, public disclosure, novelty, inventive step, enablement and sufficiency of disclosure, inventorship and entitlement, and evidence in litigation and opposition proceedings. They close with practice recommendations and argue that AI governance belongs inside a firm's risk-management system. The essay rests on Israeli and U.S. law. The Israel Patent Office (ILPO) ruled in March 2023 that AI cannot be named as an inventor, and the Tel Aviv-Yafo District Court confirmed on 31 December 2025 that "inventor" still requires human agency. The court also held that AI may be used as a tool if a human makes their own creative contribution, and that any change must come from the legislature. The essay cites Section 12 of the Israeli Patents Law on enablement and Section 39 of the 1967 Patents Law on inventorship.

The essay reports no new ruling, statute, or enforcement action. Its timing reflects the debate that followed Thaler v. Registrar of Patents, not a single triggering event. The essay does not address how Israeli courts or the ILPO would treat AI-assisted prior-art searches or drafting in a contested case, so those questions remain open. The ILPO has used AI in examination since 2023, and its June 2026 policy governing AI tools across patent, trademark, and design departments is the most concrete regulatory development the essay touches.

For practitioners, the message is to document how AI was used at each stage. A workflow that saves time can later support a novelty attack, an inventorship dispute, an entitlement challenge, or an enablement objection. Keep records of human contributions to conception, review AI outputs before filing, and check whether any AI tool exposed unpublished material. Firms should also make AI use part of their formal risk-management and governance processes.