Public's Agentic Brokerage: Users Can Deploy AI Agents That Trade Inside a Regulated Brokerage

From Stock Tips to Signed Orders

For years, retail investors have met AI at the advice layer: chatbots, screeners, and robo-advisors that suggested what to buy. Public, the U.S. fintech brokerage, is now moving that boundary. It announced that users can build AI agents that monitor markets, manage cash, and execute trades directly inside its regulated brokerage platform. The company calls this an “Agentic Brokerage.” With 4,204 likes and 580 reposts on the launch post, the market reaction suggests the idea has struck a nerve less because it is novel than because it is live.

What Public Is Actually Shipping

The product is not a sidecar research assistant or a social-trading bot. It is an agentic layer built into the brokerage itself. Users can deploy agents that watch positions, allocate cash, and place orders without the user manually confirming every click. That matters because execution, not recommendation, is where legal and operational risk concentrate. By embedding the capability inside a licensed broker-dealer, Public is making the agent an extension of the brokerage account rather than a third-party app that reads screenshots or parses emails.

The announcement lands at a moment when much of the industry is debating whether AI is overhyped. Public’s framing suggests it sees the opposite: a near-term opportunity to put autonomous software in charge of real money, not demo prompts.

Why This Is Different from the AI Trading Bots of the Past

Retail “AI trading” has long existed in a gray zone of Telegram channels, browser plugins, and unregulated signal services. Those tools typically operate outside the clearing and custody stack, which means they can alert, but they cannot settle. Public’s version collapses that gap. The agent lives inside the same regulated infrastructure that handles the trade, the cash, and the customer record.

That design choice carries implications. It means the agent can act on the account, but it also means the broker’s compliance, reporting, and risk systems can observe the agent in real time. It turns a black-box signal into an auditable workflow. Whether that supervision is deep enough remains to be seen, but the architecture is a clear departure from the arms-length AI stock bots that have dominated the conversation.

Responsibility, Risk, and the Missing Rulebook

Autonomous finance raises questions that robo-advisors did not. A robo-advisor recommends a portfolio; the customer confirms. An agent can rebalance, raise cash, and enter orders without a confirmation dialog. When an agent loses money, the distinction between advice and execution becomes the difference between a disclaimer and a lawsuit.

Regulators in the United States have focused on AI disclosure, algorithmic trading, and fiduciary duty, but they have not yet produced a clear framework for retail agents that control brokerage accounts. Public will need to show investors where the guardrails are: position limits, cash thresholds, prohibited instruments, kill switches, and clear audit trails. The company is signaling that it knows the stakes by calling its product a new category rather than a feature.

A Signal in the Bubble Noise

The launch is also notable for its timing. Headlines about AI bubbles and trillion-dollar valuations have dominated the spring, yet Public is shipping a product tied to real balances and real trades. That does not mean the market is not frothy, but it does show that some companies are moving from AI demos to financially consequential deployments. The 4,204 likes and 580 reposts are not revenue, but they are evidence that a retail audience is ready to consider handing the keys to an algorithm.

What to Watch Next

Agentic Brokerage will be measured by outcomes, not announcements. Watch three things: how regulators classify and constrain these agents, whether Public publishes explicit risk controls and audit policies, and whether the product attracts meaningful adoption beyond early adopters. If the answers are concrete, the category could expand quickly. If they are vague, the first autonomous trade may also be the last headline.