The Pulse
42AI Raises $62.5M to Predict How Traders Behave
42AI says its behavioral model uses information such as support calls and trading activity to estimate how individual traders may respond to unfamiliar market conditions. The startup says it has raised $62.5 million and is working with brok
AI.info Team · Written with AI · fact-checked ·
“Behavior is not deterministic; it’s constantly probabilistic,”
Amit Gal-Or, CEO and co-founder of 42AI, told TechCrunch
Gal-Or’s formulation is the premise behind 42AI, an Israeli startup that says it is building a model to estimate how people will act in situations they have not encountered before. TechCrunch reported that 42AI had just come out of stealth with $62.5 million in seed funding and was initially running its predictive system at financial brokerages. The system aims to forecast how individual traders may respond to volatile markets, rather than simply extend a record of their past trades.
That distinction is also the company’s central claim: people’s decisions shift with circumstances, but psychological tendencies can still shape the probabilities of what they do. 42AI says it turns those tendencies into mathematical representations, then feeds them into a transformer-based model that estimates possible decisions under different conditions. A separate language model generates explanations for the predictions.
From support calls to trading decisions
The model’s inputs can include a customer’s onboarding and identity checks, support calls, money transfers and later trading activity. 42AI CTO and co-founder Ehud Ezra told TechCrunch that the system examines signals such as sentiment, voice and how often a customer calls, alongside later choices about what and when to buy or sell. The company says it uses those varied interactions to infer behavioral traits, rather than relying only on a person’s previous actions.
42AI says it spent three years developing the approach, drawing on psychological research, neuroscience and genetics, and consulting experts in psychology and PTSD as well as Nobel laureate Dan Shechtman. Gal-Or said the team first studied thousands of people through psychological assessments and behavioral scenarios, then added government and biobank datasets covering more than 2 million people. He put the company’s customer datasets at almost 30 million people.
In live use, the system analyzes customers’ existing records, including transactions and support calls. Gal-Or said it does not need personally identifiable information and that “the model never knows who the person is.” TechCrunch noted, however, that the company still appears to create individual behavioral profiles from activity linked across customer records—a distinction that leaves questions about how those profiles are formed and governed.
What 42AI’s accuracy figures measure
42AI claims 98% accuracy when predicting market trades, but Gal-Or told TechCrunch that figure does not mean the model correctly forecasts each individual trade. It measures whether trading flows fall within a predicted range. He also said the model records an F1 score between 0.8 and 0.85 when predicting trading direction and volume.
Those qualifications matter because the numbers describe different tests: one tracks whether overall activity lands inside a range, while the F1 score measures performance across predictions of direction and volume. Ezra also acknowledged that predictions can suffer when the model lacks important context, such as a trader changing course after advice from a friend or colleague.
Gal-Or said 42AI is working with 10 large brokerages on liquidity, margin requirements and hedging exposure. The company is onboarding more customers and generating more than seven figures in revenue, though he did not specify the amount. Customers pay either a flat fee or a share of the value the platform helps create; Gal-Or said one brokerage gained about $30 million through dynamic hedging.
Funding and the advantage question
The $62.5 million accumulated over three years came from the founders and Infinity Group, Gal-Or said, so it was not all outside investment. Infinity was founded by his father, Amir Gal-Or, who is also 42AI’s chairman and a co-founder. The company plans to raise additional money.
42AI is exploring uses in gaming, fraud prevention and cybersecurity, and says it eventually wants the system to help robots and AI agents understand people. Ezra cautioned that each new domain requires training on its own rules and context. The broader ambitions therefore extend beyond the brokerage work the startup describes today.
That work raises a direct conflict: a brokerage able to anticipate customer decisions could use the insight to manage its own risk—or to gain an advantage over the people being profiled. Ezra called the concern valid and the founders pointed to regulation, monitoring and an internal ethics committee, but TechCrunch reported that they did not explain what would stop customers from using predictions against those whose behavior the system models. The company’s first market is already a setting where forecasting a client’s next move can affect how the institution manages its exposure.