Martin Lueck, co‑founder of quant pioneer AHL and president of Aspect Capital, cautioned hedge funds against surrendering trade decisions to AI. His firm, which oversees $9 billion in assets, insists on understanding the rationale behind any signal before committing capital. He warns unchecked automation could erode the firm’s accountability to clients. Lueck argues a trader must be able to attach a hypothesis to every position.

The warning arrives as many funds double‑down on machine‑learning tools. AQR founder Cliff Asness recently admitted his team lets algorithms chase patterns it cannot fully explain, noting the approach has paid off during strong market stretches but may test investor patience in downturns. Lueck recalled leaving Man Group in 1995 because the models then operated as a complete black box.

Lueck does see value in large‑language models for data wrangling, back‑testing and internal presentations, but he stresses that researchers must stay focused on the economic story, not just statistical correlations. For investors, his stance signals that firms that cannot justify algorithmic bets may face reputational risk, reinforcing demand for transparent, hypothesis‑driven quant strategies and preserve long‑term performance.