Public.com launches AI agents that monitor markets and execute trades automatically inside the app
Mar 31, 2026 with Jannick Malling
Key Points
- Public.com launches AI agents that automatically monitor markets, execute trades, and react to macroeconomic triggers like Fed rate cuts and CPI prints without user intervention.
- The platform's six-year user portfolio data gives it a structural edge over general-purpose LLMs, enabling personalized agent logic based on what customers own and the gap between stated and actual risk tolerance.
- As retail agents proliferate and close the speed gap between institutional and retail traders, stock tables and buy buttons will become obsolete within twelve months, forcing brokerages to compete on the agentic layer before larger competitors arrive.
Summary
Public.com is pitching itself as an "agentic brokerage" — a model where AI doesn't just execute trades but monitors markets, reacts to macro data, and acts on user instructions automatically. The guest frames this as a reversal of a decades-long trend: from full-service brokers who called clients with trade ideas, to discount brokers, to neo-brokers that stripped the service layer down to a button. Agents, in his view, bring that service layer back, but at a scale no human advisor could match — writing algorithmic trading scripts in seconds, running tax-loss harvesting, and analyzing risk in real time.
The product already supports macroeconomic triggers. Users can instruct the agent to move cash from a high-yield account into high-growth tech whenever the Fed cuts rates, or set rules tied to CPI prints and unemployment data. A fear-and-greed index trigger was being requested at the time of the conversation, with integration expected within days.
Public.com's structural advantage is the data it has accumulated since launching a research assistant in 2023 — roughly half the company's six-year life. It knows what users own, what they used to own, their stated risk tolerance, and, tellingly, the gap between that stated tolerance and their actual behavior. That history lets the platform build personalized agent logic that general-purpose LLMs can't replicate without the underlying portfolio and behavioral context.
On the broader market structure question, the guest expects news and macro events to get priced in faster as retail agents proliferate. Retail has grown from roughly 5% to 25% of market volume, but retail participants have historically been slow to react — not always watching screens, not always disciplined. Agents close that gap. The near-term effect may look like an alpha window for early adopters, similar to early crypto or prediction markets, before the edge compresses as adoption becomes mainstream.
The guest's sharpest prediction: in twelve months, stock tables and buy buttons will feel antique. The implied ask for investors is whether Public.com can own the agentic layer before brokerages with larger distribution, or the model providers themselves, get there first.