Israeli startup 42AI has come out of stealth with $62.5 million in seed funding, saying its behavioral world model can forecast how individuals will respond in unfamiliar situations. The company is initially running the system at a handful of financial brokerages, where it says the model can help predict how traders react to volatile conditions.
42AI describes its system as a behavioral world model. Rather than extrapolating from a person's previous actions, it tries to infer internal motivations by examining how they react to seemingly unrelated events, such as completing KYC checks, onboarding, support calls, money transfers or trading stocks. The company converts those observations into mathematical representations of traits and feeds them into a transformer-based model that lays out probabilities for decisions under different circumstances. A separate language model generates explanations for its predictions.
42AI spent three years developing the approach, combining psychological research with neuroscience and genetics, and advice from experts in psychology and PTSD and from Dan Shechtman, who was awarded the 2011 Nobel Prize in Chemistry, according to the company. Gal-Or said the team first studied thousands of people through psych assessments and behavioral scenarios before adding government and biobank datasets covering more than 2 million people. Customer datasets have since expanded to cover almost 30 million people, he said.
Ezra said the model is trained to preserve a person's underlying tendencies during simulations. "A week from now, you might be a different person, but your tendencies will always play out in certain areas to a probability or distribution," Ezra said. Gal-Or argued that "Behavior is not deterministic; it’s constantly probabilistic." He said the system does not need personally identifiable information because "the model never knows who the person is." TechCrunch reported that the company appears to create individual behavioral profiles from linked customer activity.
In live deployments, 42AI analyzes customers' existing records, including transactions and support calls. Gal-Or said 42AI is working with 10 large brokerages, which use it to manage liquidity, margin requirements and hedging exposure. The company says the model achieves 98% accuracy in predicting market trades, though Gal-Or clarified that this measures whether trading flows fall within a predicted range rather than whether each trade was predicted correctly. He said the model consistently achieves an F1 score of 0.8 to 0.85 for predicting trading direction and volume. One brokerage generated about $30 million in additional value through dynamic hedging, he claimed. Customers pay a flat fee or a percentage of the value the platform helps them gain, he said.
The $62.5 million seed funding came from the founders and Infinity Group over three years, and was not entirely raised from external investors, Gal-Or said. Infinity was founded by his father, Amir Gal-Or, who is also 42AI's chairman and a co-founder. Gal-Or said the company is generating more than "seven figures" in revenue. 42AI plans to raise more money. Beyond finance, it is exploring applications in gaming, fraud prevention and cybersecurity, and eventually wants to help robots and AI agents understand people, though Ezra acknowledged the model needs training on rules and context for each new domain.
When asked what might stop someone from using a brokerage's ability to anticipate customers' behavior to their own benefit, Ezra called it "a valid concern." The founders pointed to regulations, monitoring and an internal ethics committee. Gal-Or said he was confident the technology could help address such risks: "We are very proud that actually we are part of the solution; we are the solution to this problem."
Analysis
Why This Matters
- 42AI is selling behavioral prediction as a commercial tool for financial risk and liquidity management, not just a research idea. Its brokerage deployments will test whether firms accept that pitch.
- If the model performs as claimed, it could change how brokerages manage margin, hedging and customer risk, while raising questions about how individual customers are profiled and whether firms could use those predictions to their own advantage.
- The company's planned fundraising and its move into fraud prevention, cybersecurity and robotics will show whether investors see behavioral world models as a broader platform or a niche finance product.
Background
42AI is an Israeli startup that has spent three years developing a behavioral model. Its approach combines AI with psychology: instead of learning only from past actions, it tries to model a person's tendencies and predict a distribution of possible decisions. World models are a class of AI systems that aim to represent how an environment, or in this case a person, may evolve and respond.
The company's first market is financial brokerages, where firms already analyze customer activity for risk and compliance. Using individual behavioral profiles adds a more sensitive layer, because the system is designed to infer motivations and anticipate decisions rather than simply record what customers have done.
Key Perspectives
Amit Gal-Or, 42AI CEO: says behavior is probabilistic rather than deterministic, argues the model can predict a person's tendencies, and says the company is "part of the solution" to risks that the technology could be misused.
Ehud Ezra, 42AI CTO: says a person may change over time but underlying tendencies persist, and acknowledged that using the system to gain an advantage over profiled customers is "a valid concern."
What to Watch
- 42AI's planned new fundraising, and whether external investors back the behavioral model beyond the founders and Infinity Group.
- Results from the brokerage deployments, especially whether the claimed accuracy holds when judged outside the company's own range-based and F1 metrics.
- How the company's internal ethics committee and financial regulators treat behavioral profiling if 42AI expands into fraud prevention, cybersecurity or consumer-facing AI agents.
Checked
12 claims in this story checked against 1 source: 12 supported.
Show each claim
- Supported42AI came out of stealth with $62.5 million in seed funding.Startups | TechCrunch: “Israeli startup 42AI says it is developing a “behavioral world model” that can do just that, and it has just come out of stealth with $62.5 million in seed funding.”
- Supported42AI says its model can predict human behavior, including in unfamiliar situations.Startups | TechCrunch: “What if AI could predict what you were going to do before you decided to, even in a completely new situation? Israeli startup 42AI says it is developing a “behavioral world model” that can do just that”
- Supported42AI is initially running its system at a handful of financial brokerages to help predict how individual traders respond to volatile conditions.Startups | TechCrunch: “The company is initially running its predictive system at a handful of financial brokerages, where it says the model can help predict how individual traders will respond to volatile conditions.”
- SupportedGal-Or said, “Behavior is not deterministic; it’s constantly probabilistic.”Startups | TechCrunch: ““Behavior is not deterministic; it’s constantly probabilistic,” 42AI’s CEO and co-founder Amit Gal-Or told TechCrunch”
- SupportedEzra said, “A week from now, you might be a different person, but your tendencies will always play out in certain areas to a probability or distribution.”Startups | TechCrunch: ““A week from now, you might be a different person, but your tendencies will always play out in certain areas to a probability or distribution,” his co-founder and CTO Ehud Ezra added.”
- SupportedGal-Or said the model never knows who a person is.Startups | TechCrunch: “Gal-Or pointed out the system doesn’t need personally identifiable information, arguing that “the model never knows who the person is.””
- SupportedThe company says its model achieves 98% accuracy in predicting market trades, with that figure measuring whether trading flows fall within a predicted range rather than whether each trade was predicted correctly.Startups | TechCrunch: “In predicting market trades, 42AI claims its model achieves 98% accuracy, but Gal-Or clarified that this is more a measure of whether trading flows fall within a predicted range, not if each trade was predicted correctly.”
- SupportedGal-Or said the model achieves an F1 score of 0.8 to 0.85 for predicting trading direction and volume.Startups | TechCrunch: “He also said the model consistently achieves an F1 score of 0.8 to 0.85 for predicting trading direction and volume.”
- SupportedGal-Or said 42AI is working with 10 large brokerages that use the model to manage liquidity, margin requirements and hedging exposure.Startups | TechCrunch: “Gal-Or said 42AI is working with 10 large brokerages, who are using it to manage liquidity, margin requirements, and hedging exposure.”
- SupportedGal-Or claimed one brokerage generated about $30 million in additional value through dynamic hedging.Startups | TechCrunch: “He claimed one brokerage generated approximately $30 million in additional value through dynamic hedging.”
- SupportedEzra called the concern about someone using a brokerage’s ability to anticipate customers’ behavior to their own benefit “a valid concern.”Startups | TechCrunch: “Ezra admitted that this was “a valid concern” when I raised it”
- SupportedGal-Or said, “We are very proud that actually we are part of the solution; we are the solution to this problem.”Startups | TechCrunch: ““We are very proud that actually we are part of the solution; we are the solution to this problem,” he said.”