42AI emerges from stealth with $62.5M to build a behavioral world model

Israeli startup 42AI has officially come out of stealth, announcing $62.5 million in seed funding. The company is developing a behavioral world model designed to predict human actions and decisions even in entirely new situations.

• 42AI raised $62.5 million in seed funding from its founders and the Infinity Group.
• The technology combines psychological research, neuroscience, and data from millions of people.
• Financial brokerages are currently using the predictive system to anticipate trader behaviors.
• The startup generates seven-figure revenue and plans expansion into new domains.

How the Behavioral World Model Works

The core premise behind 42AI is that human behavior is probabilistic rather than strictly deterministic. People possess underlying psychological tendencies that influence their decisions across changing circumstances. Instead of merely extrapolating past actions, the startup measures these internal motivations through various interactions such as customer support calls, money transfers, and onboarding processes.

All of these observations are converted into mathematical representations of personal traits and fed into a transformer-based world model. A separate language model generates text explanations for the predictions. Notably, the system operates without needing personally identifiable information, mapping behavior through abstracted activity profiles.

Financial Applications and Accuracy Metrics

Currently, 42AI is working with around ten large brokerages that utilize the system to manage liquidity, margin requirements, and hedging exposure. According to the company, the model can help optimize trading flows, with one brokerage reportedly generating about $30 million in additional value through dynamic hedging.

The startup claims its model achieves 98% accuracy in terms of trading flows falling within a predicted range, alongside an F1 score of 0.8 to 0.85 for predicting trading direction and volume. However, accuracy can drop when the system misses crucial situational context, such as a trader changing decisions based on external advice.

Future Outlook and Ethical Considerations

Development of the technology took three years, drawing on expertise in psychology, PTSD research, genetics, and advice from Nobel laureate Dan Shechtman. The team initially studied thousands of participants before expanding datasets to cover nearly 30 million people.

Beyond finance, 42AI is exploring applications in gaming, fraud prevention, cybersecurity, and eventually robotics and AI agents. Addressing potential ethical concerns regarding profiling and manipulation, the founders point to industry regulations, monitoring, and an internal ethics committee to manage risks responsibly.