42AI exits stealth with $62.5 million for AI that models trader behaviour

42AI, an Israeli artificial intelligence start-up, came out of stealth on Monday, 2 November 2026, with $62.5 million in seed funding and a product it calls a "behavioral world model", software that aims to predict how people will behave. Chief executive Amit Gal-Or said the money came over three years from the founders and Infinity Group, a firm founded by his father, Amir Gal-Or, who is 42AI's chairman and a co-founder, rather than mainly from outside investors. The company says its model is already used by financial brokerages to anticipate how their clients will trade.

Most well-known AI models predict words, images or code. A "world model" tries to simulate how a system behaves, and 42AI applies the idea to people: given what traders have done before, what are they likely to do next? For a brokerage, knowing roughly how clients will trade can help it manage risk, for example by hedging, which means taking offsetting positions to limit losses when many clients bet the same way.

The way the money was raised is unusual. Seed rounds usually come from outside investors who test a company's idea before backing it. Here, most of the $62.5 million came from the founders and a family-linked firm over three years. That gives the company long runway without outside pressure, but it also means the funding figure says less about outside investors' confidence than a typical round of that size would.

The claims need careful reading. The 98% accuracy figure, as Gal-Or explained, refers to predicting a range of trading flows, not individual trades, and the $30 million gain is the company's account. Behaviour prediction also raises privacy questions. Even if the system does not need names or ID numbers, it builds profiles from linked customer activity, which regulators in many markets watch closely.

For Indian founders and investors, the launch points to a niche where AI meets financial services, a field in which Indian brokerages and fintech firms are also active. It also shows a funding route, founder and family-office capital over several years, that some deep-tech companies use before approaching venture investors.

Peepals calculation: $62.5 million spread over three years averages about $20.8 million a year (62.5 ÷ 3 ≈ 20.8). If the claimed $30 million gain at one brokerage were accurate, it would be about 48% of the company's total seed funding (30 ÷ 62.5 = 0.48). Our calculation from the reported figures.

Short-term: Coming out of stealth lets 42AI market its model openly to more brokerages and prepare for a further fundraising.

Long-term: If behaviour models prove reliable, they could spread into fraud detection, gaming and security, while drawing more attention from privacy regulators.

Who is affected: Brokerages and their retail clients; AI and fintech start-ups working on prediction; investors in deep-tech AI; and data-protection regulators.

Questions and Answers