OPENAI'S GPT-5.1 DEVELOPS GOBLIN FIXATION
OpenAI's latest models have begun obsessively referencing goblins, gremlins, and other mythical creatures in their outputs. The company has implemented prompt instructions to reduce the behavior.
OpenAI: OpenAI says its models, starting with GPT-5.1, “increasingly mentioned goblins, gremlins, and other creatures”,…
A warning for all you nerds out there.
AI Security Institute: Cybersecurity analysis: GPT-5.5 reaches a similar level of performance as Mythos Preview and is t…
OpenAI's GPT-5.5 is the second AI model capable of autonomously solving a full network attack simulation, according to t…
New results suggest Mythos' cyber threat isn't "a breakthrough specific to one model."
Taylor Lorenz / Wired: How influencers boosting US-based AI and opposing Chinese AI are paid with money tied to Leading…