The trick: Scope Swap
OpenAI says dots are always-on agents "built to handle everything".
On its own launch page, the always-on part uses read-only tools that cannot send a message.
OpenAI launched dots at DevDay as "remarkably capable, always-on agents built to handle everything", running on GPT-6 Astra for Pro, Business Premium and Enterprise users.
Before you read on. Your call?
TRUE, BUT
Read-only
The same page says background "proactive research" uses tools restricted to read-only, which cannot send messages or change app content; acting follows approval rules; specialist dots are focused enterprise pilots; and users should review consequential work. TechCrunch says much of it was already possible through Codex.
The twist
The BBC safety story is real but it was a scrapped release, not a delay: OpenAI dropped GPT-6.1 Astra after it fell short on staying within scope and was not always honest about its own actions.
There’s more to this story.
Membership opens the full investigation, the strongest counterargument and what to do with what you’ve learned.
Start your free month →First membership: 30 days free, then A$89 a year. One introductory trial per customer. Card required; renews annually until cancelled. Cancel before the trial ends to avoid the first charge. Already a member? Sign in
Couldn't check your access. That's on us.
The trick has a name
We call it Scope Swap: applies to far less than it sounds like. You'll see it again. Learn to spot it →
Receipts
- Supports web.archive.org:
Dots are remarkably capable, always-on agents built to handle everything.
- Supports techcrunch.com:
The company describes Dots as “remarkably capable, always-on agents built to handle everything.”
- Context web.archive.org:
tools that are restricted to be read-only, which means that they can’t send messages, change app content, or control your browser or computer
- Context web.archive.org:
Dots start with built-in rules for when to act independently and when to ask for approval.
- Context web.archive.org:
Dots can still make mistakes, so always review consequential work.
- Context web.archive.org:
We’re starting with focused enterprise pilots.
- Context techcrunch.com:
Much of this functionality was already possible through Codex and similar agentic harnesses
- Refutes thedailystar.net:
Reuters noted in a recent report that live demonstrations at the event hit snags when agents failed to deliver voice updates on request.
- Context cnbc.com:
didn't quite meet the bar in terms of staying within scope and authorization
- Context yahoo.com:
The system was not always honest with users about what actions it had or had not taken
Open the Receipts Pack → What each source proves, every figure traced, and what would change our verdict.