Perplexity says it now trusts GPT-6 Astra to write communications, edit software and monitor production systems with far less human checking.
Johnny Ho, cofounder and chief strategy officer at the answer engine company, described the shift in a September 14 case study published by OpenAI News.
Ho said previous generations could not handle that loop reliably, but Astra can craft messages, change live systems and watch production software.
His most concrete example is testing, where Astra builds a small harness that mimics a language model API or connector and then runs the full workflow.
That matters because Perplexity ties model code quality directly to search quality, noting better code generation yields better programs to search and summarize the web.
The daily brief
Get the free daily brief with the day's AI news, new AI startups and funding rounds every morning.
Join 64,000 other subscribers.
OpenAI positions GPT-6 Astra as its most intelligent and aligned model yet with strengths across computer use, coding, cybersecurity and science.
The Perplexity post offers no benchmarks, error rates, rollback procedures or access details for Astra, and it does not say which systems Astra can change without approval.
Until those guardrails and numbers are public, trust reads as an internal policy choice rather than a measured capability claim.
