A portfolio using Anthropic’s Claude AI model has drawn approximately twenty-seven million dollars from retail investors replicating its positions, after delivering a nineteen percent return by early August against the benchmark S&P 500 gain of twelve percent over the same timeframe. The portfolio launched in March with fifty thousand dollars under the management of finance professor Alejandro Lopez-Lira, who operates AI Finance Labs and maintains similar portfolios using competing models including ChatGPT, Grok and DeepSeek.
The Claude strategy operates with human oversight rather than autonomous execution. Lopez-Lira and his team define risk parameters and control investment decisions while using the language model for research and idea generation. By July, nearly fifty-two thousand investors had collectively allocated two hundred million dollars across seven AI-linked portfolios on the Autopilot copy-trading platform, which facilitates position replication and provides public disclosure of holdings.
The business implications for brokers centre on client engagement and product differentiation rather than proven alpha generation. Finance Magnates Intelligence identified at least ten retail brokers connecting AI agents to live accounts during the first half of this year, with Claude appearing in nine deployments. Authority levels vary considerably across implementations.
The sustainability of these results remains uncertain given the limited track record of under six months. Industry discussion at recent conferences suggests AI deployment in retail trading may serve primarily to enhance user experience and information access rather than materially improve trading outcomes.
FXnCO Insight
Brokers integrating AI trading tools should prioritize transparency around human oversight, manage performance expectations given short track records, and ensure marketing materials distinguish between AI-assisted research and autonomous decision-making to meet regulatory disclosure standards.
Source: Finance Magnates