AI agents can't yet do open-ended AI research

Factual evidence

What the source reports

Case studies from AI Snake Oil show agentic workflows fail at autonomous, open-ended research tasks despite vendor benchmarking claims.