Interactive Brokers (NASDAQ:IBKR ) founder Thomas Peterffy built his firm by automating the brokerage business. On October 16, 2026, the day after the firm reported a record pretax profit, he went on CNBC and named the problem he expects from its newest automation step, which lets clients hand trades to AI agents. “There’ll be trades taking place that the people either did not intend to do or they will. If they are losing trades, they will claim that they didn’t do it on purpose,” Peterffy said in his Squawk on the Street interview.
Interactive Brokers Just Let AI Agents Trade on Their Own
On the third quarter earnings call, management said the firm “launched autonomous AI trading, which allows eligible clients to connect the AI agent of their choice to their IBKR account and authorize it to trade autonomously.” Clients fund a dedicated section inside their existing account, and the agent can trade global stocks and options using a range of order types across day and overnight sessions.
The rollout is deliberately small. Management said approximately 3,000 clients were invited over the last several days, all of them prior users of the firm’s AI connector, which keeps a human approving each trade. Management called that connector “one of the fastest adopted features we have ever introduced” and plans to widen eligibility over the next few weeks. It expects AI adoption to lift trading volumes as clients grow more comfortable letting AI make decisions.
Losing Trades Are Where Disputes Start
Peterffy added: “It always happens that people lose money, and then they look for ways to try to recover it.”
For a broker, that is a concrete operational problem. Suppose a client authorizes an agent, the agent places a losing trade, and the client then denies it. The broker lands in the middle of a dispute over responsibility, with the client, the AI provider the client chose, and the platform that executed the order all in the frame. The dedicated section and invite-only start help here: they create a clear record of what a client authorized and limit an agent to the capital the client moves into that section. How regulators and arbitrators will treat agent-placed trades remains an open question, and Peterffy did not address it.
AI Confusion Versus What the Volume Data Shows
Peterffy also said: “There is confusion about what AI is going to bring us versus what is going to take from us.”
The firm’s own activity numbers are mixed. Interactive Brokers’ stock share volume fell 24% versus last year, which the CFO attributed mainly to low-priced stocks, where share counts run high relative to dollars traded. The company did not link that decline to AI. Total customer trades per day reached 4.3 million, up 18%, and futures activity rose 13%. Overnight trades climbed from 3.5 million in last year’s third quarter to over 10 million.
A Record Quarter Behind the Warning
Per the earnings call, commissions rose 12% to $602 million, and net interest income grew 19% to over $1.1 billion. Pretax profit set a record, and the 78% pretax margin marked the eighth consecutive quarter above 70%. Account growth ran over 30%, and the annual dividend rises from 35 cents to 60 cents per share.
Adjusted EPS of $0.67 came in just short of the $0.6716 consensus, while revenue of $1.91 billion beat the $1.86 billion estimate, according to the company’s earnings release. Management described trading as somewhat slower than in the second quarter and said there was “nothing really concerning” in the moderation.
Shares Slip a Day After the Report
As of 12:56 PM ET, shares traded at $83.29, a move of -3.18% in the session, while the stock is still up 29.91% year to date. Nothing we reviewed explains today’s decline.
What to Track as AI Trading Scales
Peterffy is right to name this weak point before it turns into a story, and the strength of the core business gives the firm room to roll out slowly. Investors should track two things. First, watch how quickly eligibility expands beyond the initial approximately 3,000 guests. Second, watch whether Interactive Brokers issues clear rules for handling disputed AI trades. If it writes those rules before its competitors do, it could set the template for the rest of the industry.