Can Active Managers Survive the AI Decade?

Active managers have long sold two products: stock‑picking skill and risk management. AI is attacking both at once.

On the research side, large language models and specialized ML tools can now ingest filings, transcripts, alternative data and news at a scale no human team can match. What used to require a floor of analysts can increasingly be done by a small team orchestrating research agents. That doesn’t automatically create alpha, but it does erase a lot of the “information processing” advantage active shops once claimed.

At the portfolio level, AI makes systematic strategies and direct indexing easier to build and distribute. A retail investor can already get factor‑tilted, tax‑optimized portfolios at ETF‑like fee levels. As AI‑native platforms improve, they’ll offer custom mandates (e.g. “low‑vol, quality, climate‑aware”) that historically required expensive separate accounts or institutional managers.

This combination compresses fees from above and below. At the top end, sophisticated allocators push active managers to embed AI or risk being out‑analyzed by peers. At the mass‑affluent level, robo‑style platforms armed with AI can deliver “good enough” portfolios at 10–30 bps. Mid‑shelf active funds charging 60–100 bps with benchmark‑like returns become very hard to justify.

Not every active manager is doomed. Firms with truly differentiated private market access, niche strategies, or deep domain expertise can use AI as leverage rather than a threat. But for the average long‑only equity or balanced fund, AI accelerates a trend that was already in motion: alpha is scarce, fees are falling, and scale or uniqueness is no longer optional.

For investors, the key is simple: investigate which managers are building AI into their process versus those still selling yesterday’s edge.