42AI Says It Accumulated $62.5 Million in Seed Funding
Its brokerage product targets client trade forecasts up to 24 hours ahead. A separate 55-scenario study tested hypothetical strategic choices.

42AI says it accumulated $62.5 million in seed funding over three years as it markets a behavioral AI system to brokerages for forecasts of client trading activity up to a day ahead. CEO and co-founder Amit Gal-Or said the money came from the company’s founders and Infinity Group. He said the total was not entirely outside capital and did not give a breakdown by contributor or a funding close date.
42AI markets its Behavioral World Model as a system for anticipating how people and AI agents might respond in volatile or unfamiliar conditions. Its brokerage product is designed to forecast each client’s trades up to 24 hours ahead by instrument, direction and volume. Gal-Or said the company was working with 10 brokerages on liquidity, margin requirements and hedging exposure.
Gal-Or cited 98% accuracy for the company’s market-trading model. He defined that figure as trading flows falling within a predicted range, rather than correctly forecasting every individual trade. He also cited an F1 score of 0.8 to 0.85 for predicting trading direction and volume. “Behavior is not deterministic; it’s constantly probabilistic,” Gal-Or said.
A 55-scenario study evaluates a different task: strategic choices, rather than brokerage trading. It assesses a Large Behavioral Model using psychometric profiles and reports 62% accuracy and a 0.45 macro-F1 score in a held-out evaluation when the model was conditioned on 20 traits. The study design set a target of 2,500 participants and used a volunteer sample.
The study cautions that responses about hypothetical or past situations may not match real-world actions. Its scenario results do not measure the brokerage system’s forecasting performance. Gal-Or did not provide a detailed evaluation method for the market-trading figures.
Gal-Or also said customer datasets had expanded to nearly 30 million people, following government and biobank datasets covering more than 2 million. He claimed dynamic hedging generated about $30 million in added value for one brokerage. He did not name the firm or provide a measurement period for that figure.