Editor's Note: This article is for informational purposes only and does not constitute financial advice. Automated investing tools involve risk and may not be suitable for all users.

Public.com has launched AI-powered investing agents for retail investors that automatically execute trades, manage cash, and monitor market conditions based on user-defined instructions in plain language. The platform, available to a first wave of users with a waitlist open for broader access, marks the debut of what Public describes as the world's first agentic brokerage.

This matters now because the model of retail investing is changing from manual order entry to continuous, AI-driven execution. Instead of placing trades directly, users define intent and allow an agent to act on their behalf as conditions are met.

The agents operate across stocks, options, crypto, treasuries, corporate bonds, and IRA accounts, running on real-time market data without requiring API keys or third-party integrations.

For retail investors and platform builders, this introduces a new way to interact with financial markets, where strategy is expressed once and execution happens continuously in the background.

In short, Public.com has moved from a click-based trading app to an intent-based investing platform — one where users express a strategy in plain language and an AI agent handles monitoring and execution.

An agentic brokerage is a trading platform that executes continuously on user-defined intent rather than waiting for manual order entry.

Public.com's AI investing agents introduce an agentic brokerage model where trades, cash management, and market monitoring are executed automatically based on user-defined strategies.

Public.com introduced its AI agent system through a dedicated Agents tab inside the Public app and a parallel web interface accessible from the main navigation. The feature was demonstrated publicly by co-CEOs Jannick Malling and Leif Abraham in a March 31 YouTube launch walkthrough, where they presented the release as the beginning of a new category in financial services: the agentic brokerage.

According to Public, an agentic brokerage is a trading platform that goes beyond order execution to act on a user's behalf. Instead of clicking buttons to place trades, users define intent — conditions, triggers, and strategies — and the platform's AI agents carry those intentions out continuously. Malling and Abraham called this a fundamental change in how retail investors interact with financial platforms. Rather than manually placing orders, the agent handles execution automatically. "What you've seen today is more than a new feature," Malling said in the launch walkthrough. "It is a fundamental shift in how you manage your portfolio."

The co-CEOs demonstrated the system live using a "buy the close, sell the open" strategy that takes advantage of overnight price action. A user types a plain-language description of the strategy, the AI asks clarifying follow-up questions — which assets, which account, how much to allocate — and then proposes a workflow viewable in a right-hand panel. Nothing goes live until the user reviews and clicks "create agent." From that point, the agent runs automatically according to the defined logic.

Agents went live for initial users on March 31. Users who do not yet see the feature in their app are directed to join a waitlist for early access.

The agent creation process on Public operates in 3 stages: prompt, refinement, and activation. A user starts by describing what they want an agent to do in plain language. The AI proposes an agent workflow based on that intent, and the user can then refine it — adjusting timing, triggers, and conditions with follow-up prompts — before activating the agent.

Once active, the agent monitors the relevant market conditions continuously and acts only when the user's defined criteria are met. The system shows both actions taken and conditions that were evaluated but not triggered, giving users a complete view of agent behavior over time. Users can access this through 3 dedicated views: the workflow itself (the agent's logic), an activity tab (a running log of every evaluation and action), and a transaction history (a ledger of all executed trades).

Public's Agent Skills currently support 4 functional categories.

The first is trading strategies, which includes equities, options, and crypto; market and limit orders; single-leg and multi-leg options; and rolling single-leg positions. Coming soon: shorting, assignment and expiration risk management, OCO and trailing stop-loss orders, and tax lot handling.

The second is indicators, which agents use to evaluate market conditions. Currently supported: Exponential Moving Average (EMA), Simple Moving Average (SMA), Relative Strength Index (RSI), Moving Average Convergence Divergence (MACD), Bollinger Bands, Average True Range (ATR), and 1-day price change.

The third is data sources, which feed real-time information into agent decisions. Currently available: real-time price feeds and VIX readings. Coming soon: ETF and index holdings, historic prices, company financials and earnings calls, dividends, bond market and futures data, and the Fear/Greed Index.

The fourth is cash management, which currently supports internal account movement and external deposits. Coming soon: portfolio performance tracking, dividend reinvesting, bank balance checks, rebalancing, and account statements.

Public also provided a set of example agent use cases on the product page to illustrate the range of the system. These include:

Public built the agent system around 3 explicit operating principles that the co-CEOs addressed directly in the launch walkthrough: transparency, security, and control.

On transparency, Public says the system is designed so there is never a "black box." Every action an agent takes — and every condition it evaluates without triggering — is visible through the Activity Feed and Transaction History. Agents operate strictly on user-defined rules, with no hidden logic.

On security, agents run on the same financial-grade infrastructure that protects standard Public accounts. Every action stays inside the user's authenticated brokerage environment. Agents on Public, the co-CEOs stated, "will never run rogue on the internet." Critically, the system requires no API keys. Because agents are built natively into the Public platform, users never have to share credentials with a third-party service to enable execution.

On control, users must approve every agent before it goes live. At any point after activation, users can edit the agent's logic, pause it, or shut it down entirely. The system also allows mid-run modifications — for example, adjusting a buy amount and seeing it instantly reflected in the agent's active workflow. In the co-CEOs' framing: "You are the ultimate decision maker here."

For decades, investing meant manually entering orders — first by phone, then through web interfaces, then through mobile apps. Public's launch makes the case that the next phase replaces order entry with intent expression. A user does not log in to place a trade; a user defines a condition and lets the agent act when that condition is met.

This matters for the AI industry specifically because it applies agentic architecture — the same model being adopted across enterprise software, healthcare, and productivity tools — directly to a regulated financial environment. Public is not wrapping a chatbot around a brokerage. The agents are natively integrated into the brokerage infrastructure, running on real-time data and executing within the same authenticated systems that handle account security. That architectural decision — native integration rather than third-party connection — is what separates Public's approach from earlier automation tools that required API bridges or external platforms.

The practical effect for retail investors is significant. Strategies that previously required either constant market monitoring or access to professional trading tools — like covered call generation, conditional hedges, or cash sweeps tied to account thresholds — can now be defined once and executed automatically. The co-CEOs highlighted covered call generation as a specific example: a user enters a target of $5,000 per month in covered call premiums, and the agent scans eligible equity positions, evaluates strike prices and expirations, and executes contracts to hit that goal.

Public says more capabilities are coming throughout 2026 and beyond, including the data sources and cash management features listed as "soon" on the product page. The platform is currently in early access, with a waitlist open for users who do not yet have agent access.

Q: What did Public.com launch and when did it go live?
A: Public.com launched AI investing agents on March 31, 2026, making them available to an initial group of users with a waitlist open for broader access. The feature introduces the ability to build automated agents that monitor markets and execute trades based on user-defined conditions, without requiring manual order entry.

Q: How do AI investing agents on Public.com actually work?
A: Users start by typing a plain-language prompt describing a strategy or task. The AI proposes an agent workflow based on that intent, asking follow-up questions to fill in specifics like which assets, which account, and how much to allocate. Users can refine the workflow before activating the agent. Once live, the agent monitors market conditions continuously and acts only when the defined criteria are met — with full visibility into both triggered actions and evaluated-but-skipped conditions through an Activity Feed and Transaction History.

Q: Why does it matter that Public built agents natively into its brokerage rather than using third-party tools?
A: Native integration means agents run on real-time market data and execute within Public's authenticated brokerage infrastructure — the same systems that protect user accounts. Users never need to generate or share API keys with an external platform. This removes both the security exposure and the latency associated with routing trades through outside services, and it allows agents to execute as soon as criteria are met rather than waiting for an external system to relay the instruction.

Q: What are the current limitations of Public.com's agent system?
A: As of the March 31 launch, several capabilities are listed as coming soon rather than currently available. These include shorting, OCO and trailing stop-loss orders, tax lot handling, assignment and expiration risk management, ETF and index holdings data, company financials and earnings call analysis, dividend reinvesting, rebalancing, and account statement access. The platform is also still in early access, meaning not all users can activate agents yet.

Q: What asset classes and account types do Public.com agents support?
A: Agents on Public support stocks, options, crypto, ETFs, treasuries, and corporate bonds. They operate across both brokerage and IRA accounts. Supported strategies include single-leg and multi-leg options, market and limit orders, rolling single-leg positions, and cash management actions like account sweeps.

Q: Do users need to know how to code to use AI investing agents on Public?
A: No. Public's agent system is built for plain-language input — users describe what they want to do, and the AI builds the workflow. No coding, no API configuration, and no third-party platform setup is required. The system is designed so that anyone with a Public account can define and activate an agent using simple prompts.

Public.com's agent launch is the first instance of agentic AI being fully integrated into a retail brokerage, not as a feature add-on, but as the core operating model of the platform.

Key point: The move from click-based order entry to intent-based investing is not a user interface improvement; it is an architectural change. Public.com is no longer just a place to buy and sell assets; it is a platform where user-defined strategies run continuously in the background, independent of whether the user is actively monitoring the market.

Who should care: AI practitioners and platform builders should pay close attention to how Public has addressed infrastructure design. By building agents natively into the brokerage rather than connecting them through APIs, Public removes latency, security risks, and credential-sharing requirements. This approach provides a model other regulated industries can evaluate.

Why this matters now: Agentic AI is expanding rapidly across industries, but financial services has been slower to adopt due to regulatory complexity and execution risk. Public's launch demonstrates that agentic systems can operate within regulated environments using approval workflows, transparent activity logs, and user-controlled safeguards.

What decision this affects: Retail investors must decide whether to continue using manual trading platforms or adopt systems that execute continuously based on defined intent. Brokerage platforms must evaluate whether traditional interfaces remain competitive as agent-driven models emerge.

In short, Public.com introduces a new operating model for retail investing, where strategy is defined once and execution happens continuously.

If the first era of online investing was defined by who could give retail traders the fastest access to markets, the next era may be defined by whose agents make the best decisions on their behalf.

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from Claude, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to Claude for assistance with research and editorial support in crafting this article.