Claims extracted
14
sentences that state something as happening or true
Uncertain
0%
0 of 14 hedged
("reportedly", "may", "according to")
Leaning
not political
takes no side on a contested political question
Publisher signals
60/100
NPR's corrections and
hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
14
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
OpenAI flags new concerning AI behavior, to track model misalignment regularly
OpenAI has disclosed six reports of "unexpected or concerning" behavior in artificial-intelligence models as the debate on AI safety becomes increasingly heated.
asserted
debate → flag → safety
The AI company also said Wednesday it was introducing a new framework for tracking, probing and disclosing AI model misalignment instances, such as new ways for the models to act without authorization, coordinate with other models or evade oversight.
asserted
models → say → oversight
OpenAI's latest announcement came as U.S. AI bosses, including OpenAI and Anthropic, are calling for a slowdown in the technology's development over safety concerns.
asserted
bosses → come → concerns
Among the new cases reported by OpenAI, an unreleased research model inserted "jailbreak-like instructions" into its own notes to disregard its normal constraints and told itself to be "freed from the roles and identities that bind other chatbots."
asserted
that → report → chatbots
In another instance, an AI "agent" uploaded files to the internet to obtain a browser citation without asking the user.
asserted
agent → upload → user
The six reports were discovered during training or evaluation over the past months, OpenAI said.
asserted
OpenAI → discover → months
"As AI systems grow more advanced and more widely deployed, we need to build a broader and better-informed consensus on the progress of alignment research," OpenAI wrote in a blog post as it disclosed the events.
asserted
it → grow → events
"Decisions about how AI development should proceed in the months and years to come need to draw on evidence that people outside the companies building frontier models can examine for themselves," the company said.
asserted
company → proceed → themselves
Wednesday's new cases followed OpenAI's disclosure in July that its rogue AI system hacked into AI startup Hugging Face.
asserted
system → follow → Face
Anthropic also said the same month that its AI models hacked into three organizations during testing.
asserted
models → say → testing
AI "agents" are becoming smarter and have become "more determined to resolve complex tasks through inter-agent collaboration, knowledge sharing, deception, and concealment," said Lian Jye Su, a chief analyst at technology research and advisory group Omdia.
asserted
Su → become → Omdia
That's making it harder to govern and contain them using traditional AI security approaches, he said.
asserted
he → make → approaches
OpenAI's new tracking and disclosure framework, meanwhile, can help push for other AI developers to also adopt similar practices.
asserted
developers → help → practices
"That said, the process remains internal and voluntary, but is a step in the right direction," Su added.
asserted
Su → say → direction
The Aporia
