What happened
A Morningstar analyst tested ChatGPT 5.2, Perplexity, Copilot, and Gemini on their ability to analyse a specific market event: the early 2026 "software selloff." During this period, companies like Experian, Expedia, Salesforce, and StubHub saw share prices drop 25-32%, while long-term US Treasury bonds rose and the overall US stock market remained flat. All tested AI models failed to identify this distinct pattern, providing garbled or contradictory explanations, and incorrectly describing some software firms as "economically sensitive" despite their recession-resistant operations.
Why it matters
The inability of leading AI models to discern novel market patterns limits their immediate utility for financial analysts and investors relying on sophisticated market intelligence. This follows recent market sentiment where AI fears have undervalued software firms, suggesting AI models struggle to interpret the very market dynamics they influence. For data scientists and CTOs evaluating AI for critical financial decision support, pattern recognition in dynamic, unstructured market events remains a significant challenge, requiring human oversight.
Subscribe for Weekly Updates
Stay ahead with our weekly AI and tech briefings, delivered every Tuesday.
