Among the things this new tool was built to do is address issues around the wording of the company’s wording of its prediction market contracts, co-founder Luana Lopes Lara said in an interview with Bloomberg on Monday (June 15).
As the report noted, bets on Kalshi often depend on the specifics of how contracts are written, and the industry has run into controversy when market phrasing hasn’t matched complex real-world events.
The report cited the example of a bet involving Netflix’s earnings call from January, and whether a company executive would say “Warner Bros.” Kalshi determined the answer was no, because the executive pronounced the name as “Warner Brothers.”
Beyond reviewing contracts, the AI agent also carries out everyday tasks like aggregating top news, analyzing competitors’ offers and making recommendations on what the exchange should list next, the report added.
“We actually have an AI engineer in the markets team, where the AI is battle-testing the entire certification — finding out if you go in this direction, maybe there’s a hole here, and all of that,” Lopes Lara said.
The news follows a report from earlier this month that Kalshi was developing an interface designed for the most highly engaged retail traders using its platform.
According to a CNBC report, the interface allows users to track popular contracts by 24-hour volumes, view all trades as they are actively placed, monitor individual contracts’ order books, and customize interfaces to see event contracts tied to their own portfolios.
The report, citing unnamed sources, said the platform could someday include research and other external information, and might extend to Kalshi’s other asset classes. Kalshi declined to comment on the report when reached by PYMNTS.
Elsewhere in the world of agentic AI, PYMNTS wrote recently about a security problem related to agentic behavior: “the web can’t verify whether an agent is authorized, by whom, or within what limits.”
A human user logs in one time and leaves a “behavioral trail” fraud detection can follow. But agents act continuously, create other agents and carry permissions that spread in ways no one could determine when access was first granted.
Research by PYMNTS Intelligence has found that 59% of firms face bot-driven fraud as an active threat and that companies lose 3.1% of annual revenue to identity gaps.
The Financial Stability Board said earlier this month it was strongly recommending that financial institutions create safeguards against agentic AI, warning that agents can produce risks that materialize quicker than human oversight can catch.
