42AI, an Israeli-American startup, came out of stealth on November 2, 2026 with $62.5 million in seed funding and what it calls a “behavioral world model” that estimates the probability that a person will make a given decision, according to TechCrunch. CEO Amit Gal-Or told TechCrunch the money came from the founders and Infinity Group over three years and was not entirely raised from external investors; the company plans to raise more. Infinity was founded by Gal-Or’s father, Amir Gal-Or, who is 42AI’s chairman and a co-founder.

How the company says it works

  • Rather than extrapolating from past actions, the model tries to measure underlying psychological tendencies by examining reactions to seemingly unrelated events, such as know-your-customer processes, onboarding, support calls, money transfers and trading, including sentiment, voice and call frequency.
  • Those observations are converted into mathematical representations of traits and fed into a transformer-based model that outputs probabilities for decisions under different circumstances. A separate language model generates explanations.
  • The team spent three years on the approach, combining psychological research with neuroscience and genetics, and drew on advice from psychology and PTSD experts and from Nobel laureate Dan Shechtman. It first studied thousands of people through assessments, then added government and biobank datasets covering more than 2 million people. Gal-Or says customer datasets now cover almost 30 million people.

Claims and limits, per TechCrunch

  • The company claims 98% accuracy on trading, but Gal-Or clarified that this measures whether trading flows fall within a predicted range, not whether each trade was predicted correctly. He said the model’s F1 score for predicting trading direction and volume is 0.8 to 0.85.
  • Gal-Or said there is “a natural ceiling” because the model lacks access to everything a person experiences, and accuracy suffers when context is missing, such as when someone changes trades on a friend’s advice.
  • Gal-Or said the system does not need personally identifiable information and “the model never knows who the person is”; TechCrunch notes the company still creates individual behavioral profiles from linked customer activity.

Customers and revenue

The model runs at a handful of financial brokerages to predict how individual traders respond to volatile conditions. Gal-Or said 42AI works with 10 large brokerages that use it to manage liquidity, margin requirements and hedging exposure. He said revenue is “seven figures” without giving a number, that customers pay a flat fee or a percentage of the value gained, and that one brokerage generated about $30 million in additional value through dynamic hedging. The company is exploring gaming, fraud prevention and cybersecurity, and eventually helping robots and AI agents understand people. Co-founder and CTO Ehud Ezra acknowledged that each new domain requires training on its own rules and context.

The misuse question

TechCrunch raises the risk that brokerages could use predictions to profit at their customers’ expense. Ezra called this “a valid concern,” and the founders pointed to regulation, monitoring and an internal ethics committee; TechCrunch says they did not explain what would stop customers from using the technology that way. Gal-Or said: “We are very proud that actually we are part of the solution; we are the solution to this problem.”

Sources and notes

This post is based on TechCrunch’s report by Ram Iyer (8:30 a.m. PST, November 2, 2026). The accuracy, customer and revenue figures are the company’s own, as attributed in the article, and were not independently verified by Lens. Lens did not contact 42AI.

Image: a candlestick price chart on a laptop screen, by Negative Space via Wikimedia Commons, licensed under CC0; cropped. Stand-in photo; it is not related to 42AI’s model and does not show the events described.