AI regulation in Israel: laws & policy (2026)

In December 2023, MIST and the Ministry of Justice jointly published Israel's first national AI policy, 'Policy, Regulation and Ethics Principles in the Field of Artificial Intelligence'. It is explicitly a strategic, non-binding document that identifies seven core risks (discrimination, human oversight, explainability, disclosure, safety, accountability, privacy) and endorses sectoral, risk-based regulation plus soft tools (sandboxes, voluntary guidance) rather than a horizontal AI Act.

Rather than a single AI law, each sectoral regulator (finance, health, transport, etc.) is responsible for applying principles to AI in its domain. An interdepartmental team's 2024 interim report developed a flexible, experimentation-friendly regulatory approach for AI in financial services, consistent with the 2023 white paper.

On 30 April 2025 the Privacy Protection Authority (PPA) issued draft guidance applying the Privacy Protection Law to AI systems, with the public comment window extended to 6 July 2025. It requires a legal basis for personal-data processing across the full AI lifecycle, limits web scraping of personal data for training to cases of implied consent, mandates defences against inference/model-extraction attacks, and extends data-subject rights to AI-generated data. If finalised it would be Israel's first authoritative AI-specific regulatory instrument.

In April 2025 the government approved a National AI Program spearheaded by the National AI Directorate within MIST. It focuses on compute (≈5,000 advanced GPUs/year 2027–2032, a 100,000-unit sovereign compute target), talent and expatriate recruitment, applied acceleration centres, and international AI partnerships. Earlier Decision 173 authorised a multi-year framework of up to NIS 500 million for 2023–2026 for AI research infrastructure, Hebrew/Arabic datasets and privacy/explainability R&D.

Israel has not enacted an EU-AI-Act-style horizontal law and such legislation is not expected within the next ~12 months; practitioners describe the pipeline as more soft regulation and sector guidance. Isolated bills (e.g. on deepfakes) have been proposed in the Knesset but none has become law.

The 2023 policy and subsequent government decisions provide for an AI Policy Coordination Centre within MIST, operated jointly with the Ministry of Justice, to consolidate expertise, issue guidance to sectoral regulators and coordinate cross-cutting AI issues — a centralising mechanism within an otherwise decentralised, sectoral model.