A Quiet Departure From Israeli Innovation Policy
On August 13, 2026, Israel's National Artificial Intelligence Directorate — housed in the Prime Minister's Office and led by Brig. Gen. (Res.) Erez Askal — published the implementation blueprint for Government Resolution 4255, the cabinet decision approved June 16, 2026 titled "Accelerating Artificial Intelligence in Israel and Establishing Global Leadership." Buried inside a plan otherwise dominated by chip procurement and quantum computing is a genuinely consequential IP policy choice: entrepreneurs who build solutions through the plan's National Institute for Impact-Driven AI Applications (NIIA) keep their intellectual property outright, conditioned only on "broad implementation" inside Israel.
That sentence is easy to skim past. It shouldn't be. Israel has run state-backed R&D funding for four decades under the Encouragement of Industrial Research and Development Law, 5744-1984 — the statute that created the body now known as the Israel Innovation Authority (IIA). Under that law's Section 19, know-how developed with state grants cannot be sold or transferred abroad without Research Committee approval, and doing so without authorization is a criminal offense carrying up to three years' imprisonment under Section 47A. Companies that do get approval typically repay grants plus interest, and ongoing commercialization triggers royalty obligations under Section 21 that continue until the original grant, with CPI-linked interest, is repaid.
The NIIA Model: Five Reactors, No Clawback
The NIIA's mechanism is five regional "acceleration reactors," each anchored to an existing center of excellence, with at least two placed outside the Tel Aviv core to spread technological employment into the periphery. Each reactor pairs a government or business lead with an implementation partner, carrying a solution from concept through national-scale rollout against nine goals spanning "optimal aging," inclusive growth, national security, productivity, and financial sovereignty. Unlike a standard IIA grant, there is no royalty stream back to the state and no export-approval gate — the entrepreneur's condition is domestic deployment, not domestic ownership.
Steelmanning the Royalty Model
The traditional IIA approach is not bureaucratic reflex; it has a real rationale. Public money is finite and taxpayer-funded, and if a founder can walk away with fully state-subsidized IP and sell it — and the underlying know-how — to a foreign acquirer the moment it's valuable, the public captures none of the upside from its own investment. Royalty clawbacks and export-approval requirements exist specifically to prevent Israeli-funded R&D from becoming, in effect, a subsidized handoff to Silicon Valley or Shenzhen. Given that Israeli cyber firms alone pulled in $3.8 billion in investment in 2024 and Israeli tech exits hit $58.8 billion in 2025, the value at stake in getting this trade-off wrong is not hypothetical.
Why the Looser Model Is Still the Right Call
Even so, the NIIA's approach is the more sensible one for the problem it's solving. Royalty administration and export-licensing regimes impose real friction — compliance overhead, valuation disputes, and a chilling effect on the kind of founder who has other options and simply won't take a grant with strings attached. A "stay operational in Israel" condition achieves the state's actual policy goal — anchoring high-value AI activity domestically — without the state trying to price and collect a return on every individual solution years down the line. It also fits Israel's broader posture on AI IP: a December 2022 Ministry of Justice opinion, publicly issued January 2023, already concluded that training AI models on copyrighted works is generally lawful in Israel under the fair use, incidental use, and transient use exceptions in the 2007 Copyright Act — making Israel one of the more development-friendly jurisdictions on AI and copyright, well before this plan existed.
Execution, Not IP Terms, Is the Real Risk
The plan unveiled last Wednesday included no details of how the budget will be subdivided.
That's the more grounded worry. The Jerusalem Post reports the 2026 budget behind the plan is roughly NIS 3 billion — a fraction of the NIS 25 billion over five years that outgoing chief scientist Prof. Ziv Nagel's earlier proposal had sought — and flags "a lack of coordination between the Directorate and the Innovation Authority," with both pursuing separate quantum-computing efforts and Israel's Finance Ministry budget department reportedly describing the overlap as the right hand not knowing what the left is doing. It is also, by that reporting, the fourth national AI plan Israel has produced in seven years, with predecessors largely shelved or underfunded.
The Bottom Line
A generous, low-friction IP policy is exactly the kind of proportionate regulatory choice this publication favors — it rewards the people who build without pretending the state can micromanage return-on-investment better than a market can. But policy design and policy delivery are separate questions. If the NIIA's reactors stand up with real money and one clear line of authority, letting founders keep clean IP is a smart bet on private capital returning where the terms don't punish success. If it becomes plan number four to stall on budget and turf disputes, the generous IP terms will be remembered as the easy part regulators got right while everything requiring sustained follow-through went unfinished.