How AI-based driving insurance could reshape car premiums, and why Israel is still behind

Car insurance remains a major financial burden for millions of Israeli drivers. An average driver pays nearly three times as much for mandatory and comprehensive coverage as peers in Germany and Britain, with prices rising even further for expensive or theft-prone models. Israeli regulators have pressured insurers to slow premium increases, and while prices have eased somewhat lately, some industry voices say that is only temporary relief, not a real correction to the sharp hikes seen since the start of the decade.

Globally, insurance is moving toward a new model built on artificial intelligence and the growing data streams of modern software-defined cars. Instead of relying only on traditional usage-based insurance, which factors in mileage, hours, roads and speed, the newer behavioral model analyzes micro-behavior, including acceleration, braking, cornering, distraction, fatigue and reaction time. It can also judge context, such as whether hard braking was caused by another driver, whether speed was appropriate for traffic or weather, and even how much a driver uses a phone or music system.

The best-known example is Tesla’s insurance program, operating in only a few U.S. states. Every Tesla collects data continuously, and the company assigns drivers a safety score from 0 to 100. Each driver starts at 90, and the next month’s premium changes based on the previous 30 days. Drivers scoring 98 to 100 can pay up to 50% less than the market average. Tesla also ignores some risky-looking events if AI concludes they were justified, and premiums can fall further when drivers use the company’s FSD autonomous driving system more often.

Other insurers are also using AI analytics, including Allstate in the United States, which can credit drivers’ accounts every few months if their driving is judged safe and consistent. The total savings on comprehensive and mandatory insurance can reach 30% to 40%, while supporters say the bigger benefit is the incentive to drive safely. In Israel, however, adoption is held back by two main obstacles: driver concern about being constantly monitored, and what the industry sees as regulator reluctance to hand so much power and data to machines. The Financial Services Authority says the opposite, arguing there are no regulatory barriers to AI-based driver analysis and that it has already approved policies with dynamic pricing tied to safety scores, as long as the scoring criteria are clear and transparent.