Interactive Brokers has introduced IBKR Autonomous Trading, allowing eligible customers to delegate order submission to connected artificial-intelligence tools across more than 75,000 listed stocks, options and exchange-traded funds worldwide. The launch moves beyond AI-generated research or draft instructions into trading that can proceed without a human approval for each individual order.
But the feature is not an unrestricted invitation for a chatbot to control an entire brokerage account. IBKR puts autonomous activity inside a separately funded AI Partition, imposes onboarding and funding rules, and lets customers disconnect a connected agent through an emergency control. Those protections are as important as the technology itself for investors assessing the risk of delegation.
Interactive Brokers announced the capability on October 21, 2026. Its investor materials and educational guidance identify material restrictions that were not prominent in some shorter accounts of the launch: access depends on the client’s jurisdiction and account eligibility, and initial funding of AI Partitions is subject to explicit dollar limits.
What Can IBKR Autonomous Trading Actually Do?
The broker’s official product announcement says clients can describe rules in natural language and allow an AI platform connected to their account to monitor markets, evaluate opportunities and submit supported orders when specified conditions are met. The platform covers more than 75,000 listed shares, options and ETFs, although that is the broad research and execution universe, not a promise that every contract, order type or exchange is enabled for every account.
Example instructions include trimming a position when its portfolio weight exceeds a threshold and selling covered calls when volatility reaches a specified level. The broker also describes alerts and conditional analyses for changes in implied volatility or falling share prices.
The major change concerns the final step in the workflow. Existing IBKR AI Integration allows an AI to formulate trade instructions but requires the customer to review and approve them. Autonomous Trading permits supported orders to be sent once the user has authorized the agent and defined operating parameters, without requesting permission again for every execution.
Milan Galik, the broker’s chief executive, presented the service as an extension of its history of automation. But machine-executed trades remain exposed to inaccurate data interpretation, ambiguous instructions, market moves and transaction costs. A platform that can trade quickly can repeat a flawed action quickly unless boundaries are enforced.
How Much Capital Can an AI Partition Control?
IBKR’s October 21 educational note gives a more precise picture of the initial safeguards. Customers allocate cash from their main brokerage account to a dedicated AI Partition rather than opening an entirely separate brokerage relationship or depositing money from an outside bank.
For eligible accounts, IBKR states that AI Partitions may collectively receive up to 10% of account equity, subject to a $1,000 minimum funding amount and a maximum of $25,000 across all such partitions. Its example allows a $2,000 account to fund a $1,000 partition, illustrating that the minimum creates a practical exception to a simplistic interpretation of the 10% percentage. All figures remain subject to product eligibility and broker restrictions.
A $100,000 account could, under the basic percentage limit, allocate as much as $10,000. A multimillion-dollar account would still face the $25,000 ceiling. These are capital allocation examples, not guaranteed maximum losses: leveraged positions, options and gap risk can produce outcomes that differ from an investor’s intended risk budget depending on which trades the platform permits.
The customer can disconnect the AI with a kill switch in Client Portal. That terminates the connection; it should not be assumed automatically to close every existing position, cancel orders already accepted by an exchange or reverse completed trades. Customers must inspect positions and order status separately after an emergency stop.
IBKR says additional guardrails, including limits on order frequency, maximum individual order sizes and minimum account-value thresholds, are planned. They should not be described as universal launch-day protections already enabled for everyone.
Who Is Eligible to Enable Trading Without Approval?
Qualification involves an initial supervised period. The broker’s guidance says users must first operate in Review & Approve mode, manually authorizing AI-generated orders. To qualify for the autonomous feature, they must complete at least 10 trades, each involving securities worth $100 or more, over a two-week period, assuming the account type, entity and jurisdiction are otherwise eligible.
IBKR indicates that a future eligibility test could potentially replace that trading-history requirement, but the proposed alternative is not the same as a presently available pathway. The broker also says its rollout is limited and expanding progressively.
On its third-quarter earnings call, management said approximately 3,000 customers had been invited into the initial autonomous trading program, mostly from users who had experience with the human-approved connector. An invitation or availability in a region does not necessarily mean every customer there can enable the feature immediately.
The supervised phase provides familiarity with how the AI translates natural-language instructions into executable orders. It is not proof that ten successful or reviewed trades establish a reliable predictive trading strategy, or that all operational risks disappear once a customer qualifies.
How Does IBKR Connect to ChatGPT, Claude and Other AI Platforms?
Interactive Brokers uses the Model Context Protocol, or MCP, as a standardized connection between accounts and external AI tools. Its July expansion announcement described compatibility with tools including ChatGPT, Claude, Grok, Perplexity and MCP-enabled developer environments.
The official integration is intended to avoid handing out brokerage passwords and raw API keys to the external AI service. This is materially different from a third-party agent logging into an ordinary brokerage website by pretending to be a person with a shared password. Nevertheless, customers still need to understand what permissions they grant and how long a session can act.
The distinction has become topical because major U.S. banks are restricting unapproved AI access to customer accounts. Interactive Brokers is pursuing the opposite route in one carefully defined setting: creating an explicit broker-controlled channel through which limited autonomous transactions can be authorized.
MCP describes a way for systems to exchange tools and information; it does not certify the investment quality of a model’s output. Nor does connecting a recognizable AI assistant mean its developer endorses a particular trade, provides investment advice or guarantees execution quality.
What New Risks Appear When the AI Submits the Order?
When a broker requires individual trade approval, a customer has a final opportunity to check the instrument, quantity, price, order type and portfolio implications. Removing that step increases speed but also moves responsibility for routine checking into software configuration, monitoring and exception controls.
Ambiguous instructions can matter. A direction to keep a stock at no more than 8% of the portfolio leaves questions about which market prices to use, whether the limit applies to all accounts, what happens after a rapid price swing and whether a trade should occur when liquidity is thin.
Options introduce additional complexity because strikes, expiries, collateral and assignment terms can produce risks that a simple natural-language request fails to specify. A covered call, for example, can cap upside and result in delivery of shares; it is not merely a way of collecting recurring income without trade-offs.
There is also potential for prompt-injection or data-quality failures when an agent reads external market commentary. IBKR’s partition and funding limitations can constrain the scope of delegated authority, but cannot guarantee that the agent chooses sensible trades.
The case contrasts with AI workflow assistants used by banks for customer servicing, where the automated action might be an account inquiry or service request rather than a security purchase or sale. Financial execution carries a direct price and market-risk consequence.
What Would Prove the Product Is Commercially Significant?
IBKR has not disclosed autonomous-trading revenue, total AI-controlled client assets, long-term strategy returns or a full breakdown of trading volumes generated by the new feature. Its initial invitation count is a rollout metric, not evidence that 3,000 autonomous strategies are live and profitable.
Investors should watch how many qualified customers fund partitions, whether the feature increases legitimate order flow, the rate of failed or cancelled orders and any publicly disclosed complaints related to unexpected execution. Risk-control reliability could matter more than short-term adoption.
Financial institutions also need to distinguish agent-enabled convenience from AI-directed attacks. A separate recent investigation documented misuse of agents to attack websites; nothing in that story establishes that IBKR’s new connector was compromised or used in that campaign.
The essential innovation is an opt-in trading channel that moves from asking an AI what to do toward permitting it to act. The equally important qualification is that IBKR has surrounded that channel with partition funding limits, initial supervised trading, eligibility requirements and a kill switch, while acknowledging that several further safeguards are still to come.
Johan Shamshad is a financial markets writer at Dave Finances covering cryptocurrencies, trading platforms, brokers, fintech, financial regulation, and developments across global markets. He previously worked at Gulf News, adding newsroom experience to his coverage of fast-moving financial and digital-asset markets.
His work focuses on identifying market-moving events, company developments, regulatory changes, product launches, and shifts in trading and financial infrastructure.
Johan contributes news and analysis designed to help readers understand not only what happened, but why a development matters and how it may affect the wider financial landscape. You can reach out to him via his social media accounts:
Linkedin: https://www.linkedin.com/in/johan-shamshad-742851262/
X: https://x.com/Yasmine_FX
Investing: https://www.investing.com/members/contributors/279781574