eToro has spent fifteen years selling the idea that investing should feel like a social network. At its Intelligence in Motion event in London this week, the company placed a much larger bet: that it should feel like talking to an agent.
The trading platform unveiled a rebuilt, AI-first mobile app anchored by Tori, a proactive AI agent, alongside a package of features that collectively push eToro from a place where you execute trades toward a place where software suggests, structures and increasingly performs them.
What’s actually new
Beyond Tori, the launch covers agent-powered portfolios; goal-based sub-accounts, which let users ringfence money against specific objectives rather than staring at one undifferentiated balance; a new desktop trading platform aimed at active traders; an expanding App Store for third-party trading tools; and instant crypto self-custody wallets powered by Zengo.
The company also refreshed its brand identity, landing on a new tagline: “Know better.” Co-founder and CEO Yoni Assia used the event to argue that AI, social investing and eToro’s community are converging into the shape retail investing will take next.
The self-custody wrinkle
The Zengo partnership is the sleeper item. Self-custody wallets let users hold their own crypto keys rather than trusting the exchange — a direct concession to the post-FTX reflex that assets you don’t control aren’t really yours. It is an unusual thing for a brokerage to build, because it makes leaving easier. That eToro shipped it anyway suggests the company reads the regulatory and cultural weather correctly.
Agentic trading is the part to watch
The App Store move is familiar platform strategy: turn a product into a surface other developers build on, and the switching costs take care of themselves. Robinhood, Revolut and a dozen neobrokers are running variations of the same play.
Agentic trading is the genuinely unsettled question. An AI agent that proactively surfaces opportunities sits somewhere between a research tool and financial advice, and that line is drawn differently in every jurisdiction eToro operates in. Regulators in the UK, EU and US have all signalled interest in how automated recommendations get disclosed, and “the model suggested it” is not a defence anyone has tested yet.
There is a behavioural risk too. eToro’s copy-trading feature already demonstrated how quickly retail investors will delegate judgment to something that looks confident. Handing that role to an agent that never sleeps, never gets bored and can act between check-ins compresses the distance between an impulse and a filled order. Whether “Know better” means the investor knows better, or the agent does, is a distinction the marketing does not resolve.
eToro hasn’t detailed which markets get the new app first, how much autonomy agent-powered portfolios actually exercise, or what guardrails sit between Tori’s suggestions and a live position. Those answers will determine whether this is a meaningful reinvention or an unusually well-staged skin. For now, it is the most complete attempt yet by a major retail broker to make the agent, not the user, the primary actor in the app.
