Forbes contributors publish independent expert analyses and insights.
A Robinhood user speaks this sentence into their phone. Seconds later, the app’s AI assistant, Cortex, has parsed the intent, set the limit price, configured the time-in-force, and staged the order for review. One tap to confirm. Done.
This is vibetrading. If it sounds familiar, it should. The same shift that transformed software development is now reshaping how retail investors interact with markets. Just as vibecoding lets developers describe what they want and let AI write the code, vibetrading lets investors express intent in plain English and let AI handle the execution.
The term itself is being popularised by Nansen, the blockchain analytics company, which this week launched what it calls “agentic trading” powered by its database of over 500 million labeled wallet addresses. “This is the most significant product launch in Nansen’s history,” said Alex Svanevik, co-founder and CEO. The company isn’t alone. Robinhood, Coinbase, and Public have all rolled out major AI trading features in recent months, each racing to make natural language the new trading interface.
The coordination is striking. Robinhood’s Cortex now lets users run technical analysis, screen stocks, and place orders through voice commands. Coinbase launched Advisor, an AI tool that builds personalized portfolios based on risk tolerance and summarizes news into actionable recommendations. Public’s “Agentic Brokerage,” announced in November, lets users create custom trading rules, “if NVIDIA drops 5%, buy it,” and have the AI execute them automatically.
The adoption numbers suggest this isn’t a gimmick. Public users created over 2,500 custom investable indices in the first month of its Generated Assets feature, compared to fewer than 4,000 ETFs that have ever existed. Nansen’s AI is now available across Solana and Base, with execution powered by partnerships with Jupiter, OKX, LI.FI, and Uniswap.
“95% of our user base is not day traders,” Jannick Malling, Public’s co-CEO, told me in an interview. “But they do have these rules, these unstructured things they want to do.” Institutional investors have executed rules-based trading for decades. Retail investors faced a gap between the idea and the execution. The agentic brokerage closes that gap.
The demand isn’t coming only from established players. Rallies.AI, a weekend project built by software engineer Hira and her husband over a few months, now monitors over $150 million in retail investments. One user connected an $8 million portfolio. The platform has fielded nearly 50,000 user questions in just months, mostly people who previously posted portfolio screenshots on Reddit asking strangers for feedback.
“Everyone notices and then everyone wants to jump in,” Hira told me. “I’m sure they are realizing that there is a good bunch of people that just ask about finance all day.”
The appeal for platforms is straightforward: vibetrading solves two problems at once.
First, it addresses what might be called the blank-slate problem. New users who open a brokerage account often don’t know what to do next. The interface is overwhelming. The options are endless. Many park cash and never trade. An AI that can suggest portfolios, explain what’s moving and why, and guide users from vague intention to executed order removes the friction that keeps deposits dormant.
Second, more activity means more volume. This isn’t cynical, it’s math. Every exchange makes money when users trade. An AI that helps users act on ideas they would otherwise abandon generates revenue the platform wouldn’t have seen. The incentives are aligned, which explains why every major exchange is scrambling to ship these features.
The AI interface itself is increasingly commoditised, the real competitive advantage is the data powering it.
Generic AI models fail at finance. Robinhood made this point in a recent presentation, noting that ChatGPT failed 97 out of 100 stock screener questions in their testing. The reason is simple: these models lack real-time market data and personalized portfolio information. They’re guessing.
Each platform’s moat is its proprietary dataset. Nansen’s 500 million labeled wallets can identify which addresses belong to smart money, tracking their movements in real time. Robinhood’s Cortex is plugged into live market feeds, analyst ratings, and each user’s portfolio down to cost basis. Coinbase’s Advisor draws on on-chain transaction history and DeFi activity. Public’s system learns from the 2,500+ custom indices its users have already created.
“We plugged it into internal systems and your account data,” a Robinhood product lead explained at a recent event. “Now Cortex knows you. It knows your portfolio, your positions, even things like your cost basis, all in real time.”
This is why the platform you choose may matter less for the chat interface and more for what the AI can see.
By the end of 2026, expect every major brokerage to offer some version of vibetrading. The technology is too compelling and the user demand too clear. Legacy platforms that don’t adapt will watch customers leave for interfaces that feel like talking to a financial advisor rather than handling a Bloomberg terminal.
The interface isn’t the product. The winners won’t be the platforms with the cleverest chatbot. They’ll be the ones with the richest and most personalized data, because in vibetrading, you’re only as good as what the AI knows.
The future of investing may be expressed in plain English. It’s built on proprietary data.
