Israel's AI English Teacher Program

TL;DR

• Israel announced a national rollout of AI-powered English instruction for all middle schools starting next school year, expanding from a 28-school pilot called Project 720.
• The program will reach 180 schools for English, maths, and science, with a parallel "English for Everyone" program targeting universal middle-school coverage. Budget: NIS 130 million (~US$45 million).
• Only 22% of Israeli ninth graders met English curriculum requirements last year. Forty percent of the country's 19,000 English teachers lack formal training.
• Pilot results won't be published until the end of July — after the national announcement but before the school-year launch.
• Teacher training begins at the start of the school year. Device procurement and infrastructure expansion are planned for the following year.
• Google and Microsoft are reportedly partners. The Education Ministry declined to confirm which tech companies are involved.

What Happened

On 11 June, Prime Minister Benjamin Netanyahu and Education Minister Yoav Kisch visited Dekel Vilnai Middle School in Ma'ale Adumim to announce that Israeli middle schoolers will begin learning English through artificial intelligence in the upcoming school year. The announcement was framed as a national milestone.

"This is one of the most significant and ambitious revolutions ever carried out in the education system," Kisch said. "It will enable every student to learn according to their own capabilities, pace of progress, and personal needs."

The initiative, called Project 720, was piloted across 28 schools. The ministry described the pilot as "a success" — but has released no data. Results are expected by the end of July, according to a ministry spokesperson who spoke to The Times of Israel. The program will expand to 180 schools next year, adding maths and science to the AI-assisted roster. A parallel program, "English for Everyone," aims to bring AI-assisted English instruction to all middle schools.

The total budget is approximately NIS 130 million (US$45 million). The ministry says the programs will initially run on schools' existing digital infrastructure. A selection process for the digital systems and tools is "currently underway." Teacher training "will begin at the start of the school year and continue throughout the year."

What It Actually Means

This is not a cautious pilot with measured rollout. It is a hard pivot — a country with a documented English-teaching crisis betting that AI can fill a gap that human teachers have not. The logic is straightforward: 40% of English teachers are untrained, only 22% of ninth graders meet curriculum standards, and the teacher shortage is structural, not cyclical. The ministry's calculation is that AI-assisted instruction, even with unresolved questions about efficacy, is better than the status quo.

But the sequence is revealing. The national announcement came before the pilot data. The teacher training begins after the program launches. The devices and infrastructure arrive the following year. The technology vendors remain undisclosed. This is not how you roll out a revolution. This is how you roll out a policy that has already been decided and is now being retrofitted with implementation details.

The program's name — "720 Degrees, representing 360 degrees around the student and 360 degrees around the teaching staff" — is marketing language. The substance is thinner. Asked about digital infrastructure, the ministry said the program "will rely on the digital infrastructure schools currently have." Asked about which tech companies are involved, the ministry declined to answer. Haaretz reported that Google and Microsoft have both been partners.

The Evidence Problem

A growing body of research complicates the ministry's framing. A 2024 Columbia University study published in PLOS ONE found that middle school students learn better from printed text than from screens. A comprehensive Brookings Institution report released earlier this year — based on interviews with over 500 students, teachers, parents, and education leaders across 50 countries, plus a review of over 400 studies — concluded that "at this point in its trajectory, the risks of utilizing AI in education overshadow its benefits."

The Brookings authors were careful: AI can be used fruitfully for learning. But they warned that the risks "undermine children's foundational development" because learning is shaped by interconnected cognitive, social, and emotional capacities — domains that screen-based AI instruction does not address.

Dr. Yishay Mor, head of the Artificial Intelligence and Education Program at Beit Berl College, told Haaretz: "If you ask a child if they prefer to learn a topic from a text or from a video, they will usually choose the video. But if you test them, you will find they remember the content much better from the text and not from the video."

An English teacher from northern Israel was blunter: "The easiest thing is to throw the kids in front of screens. That's not how you learn a language. There is no substitute for learning in front of a teacher, especially at this age. It's just another band-aid for the real problem: there are not enough teachers."

Stakeholder Landscape

Who benefits: Tech vendors (Google and Microsoft, reportedly), who gain a national-scale deployment in a country that positions itself as a global AI leader. The Education Ministry, which gets to announce a "revolution" rather than a teacher-recruitment crisis. Politicians who can claim Israel is "at the forefront of the world's leading nations in integrating AI into education."

Who is directly affected: Approximately 300,000 Israeli middle schoolers, who will receive English instruction primarily through screens. Thousands of English teachers, whose role shifts from primary instructor to — in the ministry's framing — "personal mentorship and pedagogical leadership." Whether this is an elevation or a demotion depends on implementation.

Who loses: Students who learn better through human interaction, particularly those already struggling. The 40% of English teachers without formal training, who are now being asked to facilitate AI instruction they may not be equipped to evaluate. Parents who are being asked to trust a system whose pilot results remain unpublished.

Who is watching: Education ministries worldwide. Israel is running the experiment that many countries have considered but few have attempted at national scale. If it works — by whatever metric — expect imitators. If it fails, expect it to be cited in every future debate about AI in classrooms.

What This Means for Educators and Parents

If you are an educator: the Israel case is about to become the most-cited example in global debates about AI in education. Read the Brookings report. Track the pilot data when it drops in late July. The question is not whether AI enters classrooms — it is entering — but whether the implementation sequence (data first, then rollout; training first, then teaching) is followed or inverted.

If you are a parent in Israel: ask your child's school what system they are using, who provides it, and what training teachers have received. The ministry has not disclosed the technology vendors. You have a right to know what software your child is spending hours with.

If you are a policymaker elsewhere: Israel's sequence — announce, then pilot, then train, then equip — is a case study in what not to do. The Brookings report is your starting point. The Israeli pilot data, when published, is your evidence base.

Uncertainty Ledger

• Pilot results: The ministry says they will be published by the end of July. Until then, the "success" claim is unverifiable.
• Technology vendors: Google and Microsoft are reportedly involved, but the ministry has not confirmed. The specific AI systems, their training data, their pedagogical design, and their evaluation methodology are all unknown.
• Teacher training: Scheduled to begin at the start of the school year — meaning teachers will be learning the system while students are using it.
• Infrastructure: The program launches on existing school infrastructure. Device procurement and upgrades are planned for the following year.
• Efficacy evidence: The global research base on AI-mediated language instruction for middle schoolers is thin. The Brookings report's conclusion that risks currently outweigh benefits is the most comprehensive assessment available.

Bottom Line

Israel is about to run one of the world's largest experiments in AI-mediated K–12 education, and it is doing so with the sequence inverted: announcement before evidence, deployment before training, and vendor selection before public disclosure. The teacher shortage is real. The English proficiency crisis is real. But the solution being offered — a national-scale AI deployment with unpublished pilot data and undisclosed technology partners — is a policy gamble dressed as a revolution. The pilot results, due in July, will either validate the bet or expose it. Either way, every education ministry on earth will be taking notes.

Sources: The Times of Israel (Tier 1), Haaretz (Tier 2), Brookings Institution Center for Universal Education (Tier 1), PLOS ONE / Columbia University (Tier 1), The Jerusalem Post (Tier 2)