Israeli startup 42AI says its AI predicts how traders will act, and raises $62.5 million
Ten brokerages already use its "behavioral world model" to manage risk, the company says. It acknowledged a "valid concern" that the tool could be used against the customers it profiles.
An Israeli startup called 42AI says it has built an AI that predicts how individual people will make decisions, and has $62.5 million to grow it, TechCrunch reported Monday. The company says 10 large brokerages, the firms that let people buy and sell stocks, already use it.
How it works
42AI calls its system a "behavioral world model." World models are AI systems that try to simulate how something will behave, rather than just respond to text like a chatbot. 42AI's looks for a person's lasting tendencies, such as how much risk they take, from records like support calls, onboarding paperwork, money transfers and trades. It then gives the odds of each possible decision in a new situation. CEO Amit Gal-Or said the system does not need names or other identifying details, though it still builds a profile of each customer.
Brokerages use the forecasts to manage money on hand and hedge their own risk, the company said. It claims 98% accuracy, but Gal-Or said that means trading fell within a predicted range, not that each trade was called correctly. On a stricter test, called an F1 score, it scores 0.8 to 0.85 out of 1.
Funding and concerns
The money came over three years from the founders and Infinity Group, founded by chairman Amir Gal-Or, the CEO's father, so it was not entirely outside investment.
When TechCrunch asked what stops brokerages from using the predictions against their own customers, CTO Ehud Ezra called it "a valid concern." The founders pointed to regulation and an internal ethics committee.
How it could affect markets
On its website, 42AI pitches the tool to brokerages as a way to predict each client's trades up to 24 hours ahead, by stock or other instrument, direction and size, and to "see exposure before it reaches your book," so a firm can hedge its own risk ahead of time. The uses it lists include "Alpha Strategies," a term for trying to beat the market, and fraud protection. The company told SiliconANGLE that one brokerage made about $30 million from hedging this way; that figure is the company's and could not be checked.
The U.S. Securities and Exchange Commission has said such tools cut both ways. In 2023 it said technologies that predict and guide investors' behavior "can be beneficial to investors in providing greater market access, efficiency, and returns," but that if firms use them to put their own interests first, harm could come "on a broader scale than previously possible."
Today's predictive data analytics models provide an increasing ability to make predictions about each of us as individuals. This raises possibilities that conflicts may arise to the extent that advisers or brokers are optimizing to place their interests ahead of their investors' interests.
The SEC proposed rules then that would have required brokers and advisers to eliminate or neutralize such conflicts. It withdrew the proposal in June 2025 and said any future action would need a new proposal.
For markets as a whole, the International Monetary Fund said in October 2024 that AI could make trading more efficient, deepen liquidity and improve risk management. But it warned of "increased market speed and volatility under stress, especially if trading strategies of AI models all respond to a shock in a similar manner," and suggested regulators recalibrate circuit breakers, the automatic trading pauses used during sharp drops. Gensler warned in a 2023 speech that AI could promote "herding," with market players "making similar decisions because they are getting the same signal from a base model or data aggregator."
Will it make investing a casino for ordinary people?
None of the official sources or research reviewed for this story describes investing as a casino, but regulators have warned individual investors about specific AI risks. A January 2024 alert from the SEC, FINRA, which oversees brokers, and state securities regulators said unregistered platforms were promoting AI trading systems with claims like "Our proprietary AI trading system can't lose!" It told investors to "be wary of claims — even from registered firms and professionals — that AI can guarantee amazing investment returns." It also said relying on AI chatbots for advice "could lead you to make misinformed, emotionally-driven or impulsive investment decisions."
Research from before today's AI found that frequent trading itself hurts individuals. A study of 66,465 households at a large discount broker from 1991 to 1996 by economists Brad Barber and Terrance Odean found that those who traded most earned 11.4% a year, while the market returned 17.9% and households that traded rarely earned 18.5%. "Trading is hazardous to your wealth," they wrote in the Journal of Finance in 2000.
Source: TechCrunch and SiliconANGLE, Nov. 2, 2026; 42AI website; SEC press release 2023-140 and withdrawal notice, June 12, 2025; Gary Gensler remarks, July 17, 2023; IMF Global Financial Stability Report, October 2024, Chapter 3; SEC, NASAA and FINRA investor alert, Jan. 25, 2024; Barber and Odean, "Trading Is Hazardous to Your Wealth," Journal of Finance, 2000.
- TechCrunch: 42AI comes out of stealth with $62.5M and a world model that can 'predict human behavior'
- 42AI website
- SiliconANGLE: 42AI launches with $62.5M to predict how traders act when markets turn volatile
- SEC press release 2023-140: predictive data analytics conflicts proposal (July 26, 2023)
- SEC: withdrawal of proposed predictive data analytics rules (June 12, 2025)
- SEC: Gensler remarks at the National Press Club on AI (July 17, 2023)
- IMF Global Financial Stability Report, Oct. 2024, Chapter 3 summary: AI and capital markets
- SEC/NASAA/FINRA investor alert: Artificial Intelligence and Investment Fraud (Jan. 25, 2024)
- Barber and Odean, Trading Is Hazardous to Your Wealth, Journal of Finance (2000)