AI bots: buyer beware
Market voices are warning that AI trading bots have structural limits and aren’t a retail shortcut to consistent alpha—experts urge caution and oversight (beincrypto.com). Parallel to the cautionary note, a free '7‑day AI trading' course launched this week to teach investors how to use AI for stock and portfolio evaluation rather than set‑and‑forget automation (aijourn.com).
Dragonfly Capital managing partner Haseeb Qureshi told Bankless that liability risk, market structure and the commoditized nature of current AI models are the three reasons experts won’t recommend AI trading bots. (beincrypto.com) Qureshi warned that institutional firms can scale any discovered edge to zero, arguing that firms like Jane Street could run thousands of copies of a profitable strategy—he referenced running “5,000” bots at sub‑millisecond speeds as an example. (cryptonews.net) BeInCrypto flagged that retail builders are already wiring generalist models such as Anthropic’s Claude into Polymarket and other protocols to automate trades, while no major AI lab has publicly endorsed training purpose‑built trading models for crypto. (beincrypto.com) Press materials show 7day‑ai.com launched its free seven‑day AI trading course on March 18, 2026, and identified Nate Stavseth as the platform’s chief marketing officer promoting the program. (aijourn.com) The course is described as a day‑by‑day, hands‑on curriculum that teaches using ChatGPT to assess company performance, analyze news coverage and perform portfolio evaluation, and the sign‑up process requires only an email address and phone number. (aijourn.com) 7day‑ai.com’s release says the program supplies “AI‑researched trade ideas” and touts an award claiming “Best AI Tools for Investors in the US of 2026” from BestofBestReview.com in its promotional materials. (aijourn.com) Qureshi emphasized the liability angle with concrete scenarios—he warned that a bot error could, for example, wipe out $2 million on a leveraged trade or send $10,000 to the wrong address, outcomes he said would provoke viral backlash and legal exposure. (cryptonews.net)
