--- title: "Israeli startup 42AI secures $62.5 million in seed funding" url: https://noti.group/israeli-startup-42ai-secures-62-5-million-in-seed-funding/ language: en publisher: "Noti Group" section: "Technology" published: 2026-11-02T16:30:00.000Z updated: 2026-11-02T17:43:49.850Z id: 281d6683-ef1c-4e16-b76c-f1c6e9abe192 source: "TechCrunch https://techcrunch.com/2026/11/02/42ai-comes-out-of-stealth-with-62-5m-and-a-world-model-that-can-predict-human-behavior/" attribution: "Link to https://noti.group/israeli-startup-42ai-secures-62-5-million-in-seed-funding/ and name Noti Group when you quote or summarize this story." --- # Israeli startup 42AI secures $62.5 million in seed funding The company's predictive system claims to forecast human behavior even in entirely new situations by analyzing various data points from interactions, including sentiment analysis and voice patterns. Israeli startup 42AI has emerged from stealth mode with a significant seed funding of $62.5 million, claiming to have developed a revolutionary "behavioral world model that can predict human behavior even in entirely new situations. The company's predictive system is currently being tested at several financial brokerages, where it allegedly helps forecast how individual traders will react to turbulent market conditions. According to 42AI, the model can identify underlying psychological tendencies that influence people's decisions, even when their circumstances change. CEO and co-founder Amit Gal-Or defends his company's ambitious claim by arguing that behavior is inherently probabilistic rather than deterministic. He suggests that individuals' actions are shaped by consistent psychological patterns that manifest in various areas of life. The model works by analyzing how people respond to seemingly unrelated events, such as completing Know Your Customer (KYC) processes or trading stocks. By examining these reactions, 42AI aims to uncover internal motivations rather than relying on extrapolations from past behavior. According to co-founder and CTO Ehud Ezra, the model calculates the probability of a person making a particular decision based on their underlying tendencies, which remain relatively consistent over time despite external changes. The company's world model is designed to capture an individual's unique perspective on the world by analyzing various data points from their interactions. This includes sentiment analysis of conversations with support or sales teams, voice patterns, and other behavioral metrics that provide insight into a person's tendencies. In addition to these initial observations, the model also takes into account subsequent interactions such as purchasing habits and market trends. By combining these data points, the company aims to create a comprehensive understanding of how an individual perceives and navigates their environment. The world model then uses mathematical representations of these traits to lay out probabilities for possible decisions under different circumstances. This is achieved through a transformer-based architecture that allows the model to process complex patterns and relationships between variables. To ensure the accuracy of its predictions, the company's approach focuses on preserving an individual's underlying tendencies during simulations. This means that if a person consistently exhibits risk-averse behavior in real-world situations, the model will not suddenly treat them as a risk-taker. The development of this approach involved combining insights from psychology, neuroscience, and genetics with expert advice from professionals in these fields, including Dan Shechtman, who was awarded the 2011 Nobel Prize in Chemistry. This multidisciplinary approach has been ongoing for three years, resulting in a sophisticated model capable of predicting human behavior. The model's training data has expanded significantly since its inception, incorporating thousands of people's psych assessments and behavioral scenarios, as well as government and biobank datasets covering over 2 million individuals. This vast dataset is complemented by the company's own customer records, which now span nearly 30 million people. In live deployments, the model analyzes customers' existing records, including transactions and support calls, to make predictions about their behavior. The system does not require personally identifiable information, with Gal-Or stating that the model never knows who the person is. However, it appears that individual behavioral profiles are still created from linked customer activity. In the context of market trades, 42AI claims its model achieves an impressive 98% accuracy. However, this figure refers to whether trading flows fall within a predicted range, rather than each trade being correctly predicted. The company's model also consistently achieves an F1 score of 0.8 to 0.85 for predicting trading direction and volume. This metric is used in machine learning to balance correct predictions against false alarms and missed events, providing a nuanced understanding of the system's performance. 42AI is currently working with 10 large brokerages that are utilizing its platform to manage liquidity, margin requirements, and hedging exposure. The company is in the process of onboarding more customers and has reportedly generated significant revenue, exceeding seven figures. Customers pay a flat fee or a percentage of the value gained through the use of 42AI's platform. One brokerage allegedly generated approximately $30 million in additional value through dynamic hedging using the startup's technology. The seed funding for 42AI came from its founders and Infinity Group over a period of three years, with some portion not coming from external investors. The company plans to raise more money in the future. In addition to finance, 42AI is exploring applications for its model in gaming, fraud prevention and cybersecurity. The goal is to eventually use the platform to enable robots and AI agents to understand human behavior, but this may require significant training on new rules and context. The potential implications of 42AI's technology are being closely examined by its founders. One concern raised was that if a brokerage can anticipate customer behavior, what's to stop someone from using that knowledge for their own benefit? Regulations, monitoring, and an internal ethics committee have been cited as measures to mitigate this risk, but the exact mechanisms in place to prevent customers from exploiting the technology remain unclear. The company's technology has been touted as a solution for mitigating risks associated with its potential use in malicious activities. Gal-Or expressed confidence that the tech can help address these concerns, stating that 42AI is the solution to this problem." A key aspect of the company's approach is its focus on developing a world model capable of predicting human behavior. This technology has sparked both excitement and concern among experts. Some have raised questions about how 42AI plans to prevent customers from exploiting the tech for malicious purposes. The company's world model has been trained on a vast dataset and can reportedly predict human behavior with high accuracy. According to its creators, this technology could have significant applications in fields such as marketing and advertising. A key concern surrounding the use of this technology is how it will be regulated and monitored to prevent malicious exploitation by customers. Despite these questions, 42AI's founders are optimistic about the potential benefits of their model, which they believe can revolutionize various industries with its predictive capabilities. --- Source: [TechCrunch](https://techcrunch.com/2026/11/02/42ai-comes-out-of-stealth-with-62-5m-and-a-world-model-that-can-predict-human-behavior/) Published by Noti Group: https://noti.group/israeli-startup-42ai-secures-62-5-million-in-seed-funding/