Robinhood's AI trading agents and HOOD Summit 2026 product launch

More than 1,500 retail traders, industry executives, and market thinkers attended HOOD Summit 2026 in Houston on September 29, where Robinhood unveiled 16 major product updates headlined by Robinhood Agents ^[12], a fully embedded AI experience that can analyze markets, build strategies, and trade on a customer's behalf around the clock ^[1]. Inside the app, customers choose a model from OpenAI or Anthropic, trade approval is on by default, and no margin is available at launch ^[2]. In its own keynote, Robinhood said GPT-6 "Luna" would be free to use through the end of the year via a direct partnership with OpenAI, while other models are billed per-token at standard lab rates. A forthcoming "Loops" feature turns a one-off prompted strategy into a standing instruction the agent carries out on repeat, day and night ^[3], pitched in the keynote as being "like having a team of hedge fund analysts working for you." Robinhood Agents is only one piece of a broader push: the company plans 24/7 weekend trading in a curated set of U.S. equities and ETFs via its own alternative trading system, Bruce ATS, pending regulatory review and targeted for early 2027 ^[1][4]; crypto perpetual futures on eight assets through Bitstamp, with 10x leverage on bitcoin and ether and 3x on six other listed tokens - SOL, XRP, DOGE, ADA, LINK, and HYPE ^[3][4]; earnings prediction contracts tied to company-specific KPIs via Cboe, fee-free through year-end with a $1.00 maximum payout per contract ^[1]; and new OCO orders for equities, a 35% expansion of options trading hours, and intraday margin of up to 4x buying power during regular hours ^[5]. Robinhood Agents also gained Agent Apps, eleven third-party data and tool integrations such as Unusual Whales Options Trader and Nasdaq Investor Intelligence, each bundled with a one-month free trial ^[1], and Robinhood Social - which lets users follow investors and share trades - came out of beta for all eligible U.S. customers ^[6].

Despite the "trade on your behalf" framing, the rollout is narrower than the headlines suggest. Trade approval is on by default and no margin is available at launch ^[2], and reporting built around a companion Forbes interview adds further texture: agentic accounts are separate, dedicated accounts isolated from a customer's standard brokerage account, users can pause autonomous trading at any time and preview trades before or after execution, and the beta covers equities only - options, crypto, event contracts, and futures traded through agents are described as "coming in later releases." That scoping lines up with an early user's account of actually using the product, who reported that it currently only recommends trades and still requires manual confirmation before anything executes - well short of the "fully autonomous" framing in some press coverage. Robinhood nonetheless reports strong uptake since the original Agentic Trading soft launch in May 2026 ^[11]: more than 150,000 customers have opened agentic trading accounts, and agents now use Robinhood's tools almost 30 million times a day ^[2][3].

Sell-side reaction leaned positive. Morgan Stanley's Michael Cyprys said Robinhood delivered exactly the two products he anticipated most - Agents and U.S. perpetual futures - and maintained his Buy rating and $150 price target ^[9]. Deutsche Bank's Brian Bedell called the announcements evidence of Robinhood's aggressive expansion across AI, prediction markets, and cryptocurrency, even as he trimmed his price target slightly, from $138 to $134 ^[9]. The market responded in kind: HOOD shares rose roughly 4.7-5% in the hours after the summit ^[2][7]. For context, Robinhood is explicitly chasing share from Coinbase and Charles Schwab, the latter of which had already added 24/7 crypto futures trading on its thinkorswim platform earlier in 2026 ^[7][9].

The product's risk disclosures are blunt: customers bear full responsibility for trades their AI agent executes and for any use of their data by third-party LLM providers, and Robinhood says it does not control, supervise, monitor, recommend, or audit the agents themselves ^[3]. The Loops feature compounds this, since it can place, modify, or cancel trades automatically without prompting for approval, including while a customer is asleep or away from the app ^[3]. That autonomy is exactly what worries regulators: Bank of England Deputy Governor Sarah Breeden warned that autonomous AI trading agents acting at scale could amplify volatility during periods of market stress ^[3], and FINRA's 2026 oversight report flagged a need for novel supervision frameworks built specifically around autonomous agents ^[10]. Separately, researchers from Wharton and the Hong Kong University of Science and Technology found in simulation that AI trading agents can sustain above-market profits even without any explicit communication, agreement, or coordination between them ^[3]- a dynamic regulators would likely read as a new channel for market concentration even absent outright collusion.

Robinhood's own promotion of the launch was unreservedly upbeat, framing Agents as something users can "spin up... in under a minute" with "safety we built in from the start." Not everyone is convinced the strain shows up where Robinhood is looking: one fintech infrastructure commentator argued that mass agentic trading volume shifts the real bottleneck from the interface to the settlement layer, since thousands of agents placing and canceling orders in milliseconds stresses trade-settlement infrastructure rather than just the mobile app's UI - a concern distinct from the volatility-at-scale warnings coming from regulators. CEO Vlad Tenev has himself acknowledged a related, narrower version of the same crowding problem: if every customer tells an identical agent to "go out and make me money," that strategy stops being the best one simply because everyone is running it ^[8]. Underneath the product debate sits an older trust problem: retail discussion of the launch skewed skeptical, rooted less in the technology than in Robinhood's reputation coming out of the GameStop era and lingering suspicion about its payment-for-order-flow business model, with some commenters speculating Robinhood profits regardless of whether a given agent's trades win or lose.