Public launches prediction markets and AI agents in Kalshi partnership
Public launched prediction markets and AI agents, with a Kalshi tie-up, putting automated market analysis and trading workflows in focus for investors.
Public has launched prediction markets and introduced AI agents designed for use in them, according to a company announcement reported by PR Newswire. Fortune separately reported that the investing platform is tying the offering to Kalshi, while CNBC interviewed Public co-founder and co-CEO Leif Abraham about the launch and the company’s broader integration of AI agents.
The move brings automated AI tools into a category where users trade contracts tied to the likelihood of events. It also places Public alongside a growing group of financial platforms experimenting with software that can interpret information, surface potential trades or assist with execution. The available source material confirms the launch and the Kalshi relationship, but does not provide enough detail to establish how autonomous the agents are, which markets they cover, or whether they can place trades without user approval.
Public connects AI agents to prediction markets
PR Newswire’s announcement identifies the new product as “AI Agents for Prediction Markets.” Fortune’s coverage describes the initiative as bringing “AI trading agents” to prediction markets through a Kalshi tie-up. Taken together, those reports indicate that Public is not merely adding a general-purpose chatbot to its investing app. The company is positioning AI specifically around the research and trading workflow for event-based contracts.
That distinction matters for product teams. An agent operating in prediction markets may need to monitor news, interpret probabilities, compare contract prices and explain the assumptions behind a position. Those tasks differ from the portfolio summaries and financial education features commonly associated with consumer investing assistants.
The evidence does not establish whether Public built the underlying models itself, which model providers are involved, or what data sources the agents use. It also does not specify whether the agents generate recommendations, prepare orders for review, or execute trades directly. Those unanswered questions are central to assessing the product’s risk and practical usefulness.
Kalshi provides the market connection
Fortune’s headline identifies Kalshi as Public’s partner for the initiative. Kalshi operates an event-contract marketplace, making the relationship a potentially important piece of the product’s structure. Public can use the partnership to connect its audience to prediction-market activity while adding an AI layer around discovery and decision-making.
The source evidence does not include the commercial terms of the partnership, the jurisdictions in which the offering is available, or the specific contracts that Public customers can access. It is therefore too early to characterize the arrangement as a broad distribution deal or to assess how much activity it may generate for either company.
For Kalshi, a relationship with a consumer investing platform could expand the reach of event-based markets beyond users who already seek out specialized prediction-market products. For Public, the partnership may provide a way to add a new asset or market category without operating every underlying marketplace function itself. Whether that model produces sustained engagement will depend on liquidity, contract selection, user protections and the quality of the AI assistance.
What the launch says about AI-assisted trading
The launch reflects a specific direction in financial software: using AI agents to turn large volumes of public information into actions that fit an investment workflow. In prediction markets, the underlying question is often expressed as a probability rather than a conventional share price. An AI system could help users organize evidence and understand how changing information affects that probability.
However, a fluent explanation is not the same as a reliable forecast. Event markets can move quickly on incomplete information, and an agent may misread a source, miss a settlement rule or give excessive weight to a recent headline. If an agent is connected to order placement, those errors can become financial losses rather than merely poor advice.
That makes transparency especially important. Builders will want to know what information an agent used, when it retrieved it, whether it distinguishes facts from forecasts and how it handles uncertainty. Enterprise buyers and regulated financial firms will also examine audit logs, approval controls, suitability checks and the separation between an AI-generated suggestion and a completed order.
Neither the PR Newswire announcement nor the cited media summaries provide those technical or governance details. Public’s launch should therefore be understood as a product announcement, not evidence that AI agents can consistently outperform human traders or prediction-market benchmarks.
Evidence behind the announcement
The strongest confirmed facts in the available reporting are limited. PR Newswire reports that Public launched AI agents for prediction markets. Fortune reports a Kalshi tie-up and frames the agents as trading tools. CNBC’s coverage centers on comments from Leif Abraham about the launch of prediction markets and the integration of AI agents.
The supplied source material contains no full article text, direct quotations, performance results, customer counts, trading volumes or independent evaluations. It also contains no evidence of adoption beyond the fact that the launch was covered by the three outlets. Any claims about improved returns, user growth or agent accuracy would need additional documentation from Public, Kalshi or independent researchers.
This distinction is important because financial AI announcements often combine a confirmed product release with longer-term expectations about automation. In this case, the reporting supports the existence of the offering and the partnership, but not a conclusion about effectiveness or scale.
Implications for builders and investors
For AI builders, Public’s move raises the bar for agents that operate in high-consequence workflows. A useful system will need more than retrieval and natural-language output. It must handle market-specific rules, expose the reasoning and evidence behind a probability estimate, prevent unauthorized actions and give users a clear way to challenge or correct its assumptions.
For product teams, the central design choice will be the boundary between assistance and execution. A research agent that summarizes competing evidence carries different risks from an agent that submits an order. Human confirmation, spending limits and clear disclosures may be essential if Public allows the system to move beyond analysis.
For investors, the launch creates another way to access prediction markets but does not remove the underlying risks. Users will still need to understand contract settlement, liquidity, fees and the possibility that an AI-generated view is wrong. The presence of an agent may make complex markets easier to navigate, but it can also make trading feel more certain than the evidence warrants.
What to watch next
The next signals will be practical rather than promotional. Public’s product documentation should clarify whether its AI agents only analyze markets or can place trades, which models and data sources support them, and what review controls apply. Availability by country and account type will show how widely the service can be used.
Market observers should also watch for details on the Kalshi integration, including eligible contracts, order routing, liquidity and settlement responsibilities. Independent analysis of agent recommendations, error rates and user outcomes would provide a stronger test than launch-day claims. Finally, disclosures about incidents, model limitations and regulatory treatment will indicate how Public plans to manage the risks of combining AI agents with financial transactions.
Creati.ai perspective
Public’s launch is notable because it connects AI agents to a live decision workflow rather than presenting them as a standalone conversational feature. The Kalshi relationship gives the announcement a concrete market layer, but the product’s significance will depend on controls and evidence that are not yet available in the supplied reporting.
The key question is not whether an agent can produce a plausible market thesis. It is whether users can understand, verify and safely act on that thesis. Public will need to demonstrate that distinction as prediction markets and AI-assisted trading become more accessible.
