For two years, invisible Unicode characters have been the neat trick of AI security research. You hide instructions inside them. A person sees nothing. A language model reads them and does as it is told. Microsoft has now found somebody using the same …
OpenAI acknowledged its role in a recently reported incident where AI agents took over a German wiki forum.
As OpenAI unveils GPT-6 Astra, cybersecurity experts question whether the model’s ‘recurrent depth’ reasoning was properly tested.
Plus: Tens of millions of US and Canadian drivers’ licenses go up for sale on the dark web, the US military finally tries to tackle the risk online ad data poses to troops, and more.
Google did not share the details about the bug being actively used.
GPT-6 Astra was paused a month ago for triggering safety protocols and now it’s being slowly rolled out.
OpenAI released GPT-6 Astra on 3 September, saying it outperforms every rival including Anthropic’s Claude and Google’s Gemini. Astra is “state-of-the-art on computer use, browsing, software engineering, cyber security, science, and professional work”,…
After hackers failed to extort Manchester Airports Group and posted the data of nearly 9 million people online, we asked the experts what’s really going on.
A total of eight flaws were fixed, none of which were exploited in the wild.
A swarm of rogue AI agents from OpenAI reportedly commandeered a German website and transformed it into a messaging board for other agents, with officials staying quiet about the incident for weeks as the company prepared to launch its most advanced model yet, Astra. The finding adds to intensifying concern surrounding oversight at frontier AI […]