29 Million Accounts Meet the Machines: Robinhood Agents Brings In-App AI Trading to Every Customer

Robinhood has just crossed a line that Wall Street has been eyeing nervously for two years. At its HOOD Summit 2026 event in Houston, the company announced that Robinhood Agents — AI trading agents that live entirely inside the Robinhood app — are rolling out to all of its roughly 29 million customers, starting this week. No API keys, no MCP configuration, no coding. A user picks a model from several leading AI labs (OpenAI’s GPT-6 Luna and GPT-6 Sol, or Anthropic’s Opus 4.8), gives the agent a name, opens a dedicated agentic account, and starts issuing plain-English instructions like “Buy $200 of Ford stock” — or considerably more elaborate marching orders.

It is the first time a major brokerage has shipped a nontechnical trading agent to a mass-market customer base, and it lands alongside a bundle of other boundary-pushing announcements: Agent Apps (premium third-party data tools for agents), Loops (standing automated strategies), 24/7 weekend equities trading slated for early next year, perpetual futures with up to 10x leverage on BTC and ETH, and Cboe-listed earnings contracts on company KPIs.

From developer experiment to default experience

The May release was the proof of concept. Robinhood launched its Trading MCP and Agentic Trading accounts on May 27, 2026, giving technically inclined users a way to wire their own third-party agents into the brokerage. The adoption numbers the company disclosed this week turned that experiment into a mandate: over 150,000 customers have opened agentic trading accounts, and agents now invoke Robinhood’s tools almost 30 million times per day. That is machine traffic, not human traffic — Robinhood’s platform is already, by volume of tool calls, predominantly agentic.

Robinhood Agents closes the loop for everyone else. “The experience now lives right inside the Robinhood app, meaning anyone, regardless of their familiarity with AI, can easily build their own agents to help them develop powerful strategies, run detailed research, or trade automatically,” the company wrote in its newsroom post. CEO Vlad Tenev framed the whole HOOD Summit lineup as a deliberate transfer of power: “We’re making Robinhood the best place in the world for active traders by delivering tools once reserved for hedge funds, big banks, and quant firms.”

What an agent actually does

The mechanics matter, because they sketch what “agentic finance” means for ordinary users:

• Dedicated accounts. An agent can only touch funds in its own agentic trading account, strictly separated from the user’s main portfolio.
• Manual approval by default. A confirmation mode requiring human sign-off on each trade is shown during setup and defaults to on, adjustable at any time (crypto trade approvals must be enabled in CT, NY, and CA).
• Loops. Coming soon, users will be able to convert a strategy into a standing instruction the agent executes on repeat — “check the market every morning and execute a trade when certain conditions are met, or run a continuous overnight strategy to look for opportunities while you sleep.”
• Agent Apps. Eleven third-party providers at launch or shortly after — options-flow analytics from Unusual Whales, congressional-trading trackers from Quiver Quantitative, on-chain analytics from Token Terminal, satellite-imagery feeds, supply-chain forecasting, social-narrative detectors — subscription-priced between $5 and $30 a month, each with a one-month free trial. The pitch is explicit: retail agents get institutional-grade data.
• Free model usage. Usage on OpenAI’s GPT-6 Luna is free until the end of the year; OpenAI and Anthropic agents otherwise bill at standard token rates.

Abhishek Fatehpuria, VP of Product Management, put the safety framing front and center: “With safety and security at its core, we’re setting the standard for what agentic finance can be.”

The market structure underneath

Two of the accompanying announcements make the agent rollout more potent. First, weekend equities trading — pending regulatory review, a curated list of US stocks and ETFs will trade 24/7 including weekends via Bruce ATS, extending the Robinhood 24 Hour Market that has run Sunday 8pm ET to Friday 8pm ET since 2023. An overnight Loop that hunts opportunities while you sleep is only useful if the market is actually open while you sleep.

Second, perpetual futures on BTC, ETH, SOL, XRP, DOGE, ADA, LINK, and HYPE with no expiration dates — up to 10x leverage on BTC and ETH perpetuals, 3x on the rest — offered through Robinhood Derivatives via Bitstamp at one basis point per trade through year-end. Plus OCO orders for equities in October, 4x intraday buying power for eligible margin customers, and options hours extended to 7:30am–4:15pm ET.

Why this matters beyond one brokerage

Fortune’s assessment is hard to improve on: Robinhood is the first to release a nontechnical trading agent for a giant customer base, “a step that has the potential to alter how Americans invest, and that could have ripple effects on markets more broadly.” The open questions it raises are the right ones:

• Herding and volatility. If agents confer with one another — or simply converge on similar signals from similar data feeds — they could move en masse into or out of a given asset. Thirty million tool calls a day is already a systemic-scale behavior pattern; what happens at a hundred million?
• Liability vacuum. Robinhood’s position is that hosting agents is not providing financial advice, and that agent suggestions are akin to “asking the internet or a friend.” Whether regulators and courts accept that analogy for an entity that executes trades is an entirely unsettled question.
• Cost opacity. Executives say per-transaction agent costs will be negligible, but token-metered pricing is denominated in compute, not in trades. Research-heavy strategies on premium data feeds could add up in ways retail users don’t model.
• Competitive cascade. eToro, Public, and Coinbase already let users connect agents via MCP tooling. With Robinhood moving agents into the default app experience, mainstream incumbents like Schwab and Fidelity now face the classic disruptor’s dilemma — and the agent platform race arrives at retail.

For now, the empirical signals point one direction: 150,000 self-selected power users became 29 million eligible ones overnight. Fortune sketches the endpoint plainly — “an environment where agents are placing billions of trades a day, and where ordinary investors are deploying elaborate strategies in the new corners of the market. How this affects market performance and wealth accumulation remains to be seen.”

The democratization arc that started with zero-commission trades in 2013 has reached its strangest chapter: the machine making the trade is now the product.