The China Securities Association told brokerages in a 2025 securities research
operations report that the industry lacks clear boundaries and controls for AI
use in research and that some analysts have publicly used unverified,
AI‑generated errors, damaging the sector’s credibility. The report calls for
internal control and content‑review mechanisms to vet AI outputs for accuracy,
logic and compliance, and urges a ban on analysts directly relying on unverified
AI‑generated data, views, market simulations and research conclusions to guard
against model hallucination risk.
