FrontierThe story, in brief
OpenAI claims ChatGPT’s new default model hallucinates way less
52.5% fewer hallucinations. OpenAI's new GPT-5.5 Instant claims a major leap on high-stakes domains.
The KeyNews take
Why it matters
OpenAI is addressing hallucination—a foundational reliability gap that has blocked enterprise adoption in medicine, law, and finance. Measurable factuality improvements at scale signal models are moving from experimental to deployable.
The key facts
6 to knowGPT-5.5 Instant model launched as new ChatGPT default
52.5% reduction in hallucinated claims vs GPT-5.3 Instant on high-stakes prompts
37.3% reduction in inaccurate claims on flagged challenging conversations
High-stakes domains tested: medicine, law, finance
Internal evaluation-based claims
Published: May 5, 2026
Go to the source
The Verge AItheverge.com
Publisher excerpt: OpenAI's newest default model for ChatGPT might not make stuff up as much. Hallucinations have been an ongoing problem for AI models, but OpenAI says its new GPT-5.5 Instant model has "significant improvements in factuality across the board." The company claims that, based on "internal…