42AI emerges with $62.5M to predict individual traders’ behavior for their brokers

Israeli startup 42AI has emerged with $62.5 million in seed funding and what it calls a behavioral world model: a system estimating how a specific person is likely to decide, even in situations they haven’t faced. It already runs at brokerages, which the company says use it to manage liquidity, margin requirements and hedging. The founders told TechCrunch the model doesn’t need to know who a person is, but it builds individual profiles from linked activity like support calls, transfers and trading history. That puts a profile of each customer’s tendencies in the hands of the firm on the other side of their trades. It resembles the ambitions of Mirror Particle (Mirror Particle).

Modeling stable tendencies rather than extrapolating recent clicks is an interesting bet, and a brokerage that anticipates panic selling can manage risk more safely and be a steadier counterparty. But a broker that can predict a customer’s next move under stress has an obvious temptation to profit from it. When TechCrunch raised this, the co-founder called it a valid concern; the company pointed to regulation, monitoring and an internal ethics committee without explaining what would stop clients from using profiles against the people profiled. An ethics committee isn’t an answer at this scale, echoing debates over who controls AI’s safeguards (emergency brake).

The model reads voice and sentiment from support and sales calls, and customer datasets now cover almost 30 million people, after early work drawing on government and biobank data covering more than 2 million. Most never signed up to be modeled this way. The company spent three years on the approach, drawing on psychology, neuroscience and genetics, with advice including Nobel laureate Dan Shechtman; traits become mathematical representations for a transformer, and a separate language model explains predictions. Founders admit accuracy drops without situational context, such as a trader acting on a friend’s tip. If it spreads into gaming, fraud prevention, cybersecurity and agents as hoped, disclosure to the people modeled, use limits and early regulatory questions are needed. More in our funding coverage and finance coverage.