“ChatGPT | Research, synthesis and explaining complex developments | Can make factual mistakes despite convincing explanations”
“Research, synthesis and explaining complex developments”
“ChatGPT is a strong choice for synthesising multiple reports, explaining the background and identifying the implications of an event.”
“Can make factual mistakes despite convincing explanations”
“Claude | Careful analysis, nuanced explanations and source evaluation | Can miss or misinterpret newly emerging information”
“Careful analysis, nuanced explanations and source evaluation”
“Claude is worth considering for examining arguments, comparing interpretations and analysing lengthy source material.”
“Can miss or misinterpret newly emerging information”
“Strong real-time news retrieval and factual question answering”
“Gemini performed particularly well in Stanford's same-day news accuracy test.”
“Source attribution and consistency have been concerns”
“Perplexity | Finding current reporting and linking directly to sources | Source quality doesn't automatically guarantee accurate conclusions”
“Finding current reporting and linking directly to sources”
“Perplexity performed particularly well on source credibility”
“Source quality doesn't automatically guarantee accurate conclusions”
“Microsoft Copilot | Web-assisted answers and integration with Microsoft's ecosystem | Can struggle with sourcing and nuanced reporting”
“Web-assisted answers and integration with Microsoft's ecosystem”
“Can struggle with sourcing and nuanced reporting”
“Rapid access to social-media discussions and breaking developments”
“Grok 4 retained 70% accuracy”
“Grok 4 | 95.0%”
“Needs careful verification of claims circulating on social media”
“GPT-5 scored just 19% in that particular test”
“GPT-5 scored just 19% in that particular test”