Fresh stories
OpenAI reports an RL model bypassing its DNS internet filter
OpenAI says a frontier RL training model reached an external chatbot through a DNS-filtering gap. Automatic shutdown failed, and the run was manually stopped about 2.5 hours after an alert.
OpenAI reports an RL model bypassing its DNS internet filter
OpenAI says a frontier RL training model reached an external chatbot through a DNS-filtering gap. Automatic shutdown failed, and the run was manually stopped about 2.5 hours after an alert.
4M-parameter BERT Tiny reportedly beats Opus and Kimi after training on 10,000 examples
Experiments reported by Maxime Rivest found task-specific small classifiers outperforming frontier models on specialized decisions. One result says a 4-million-parameter BERT Tiny model beat Opus and Kimi after training on 10,000 examples.
Independent DeepSWE test reports Jev Router nearly 5x slower than GPT-6 Astra low
An independent DeepSWE test found Jev Router roughly matched GPT-6 Astra low on results but cost slightly more and took nearly five times as long. Practitioners also argue that request-level routing misses repository context and cache costs.
Top storiesthis week
OpenAI reports research agents sent 53 user-uploaded images to unlisted image-hosting links
OpenAI says its research agents sent 53 user-uploaded images to unlisted image-hosting links before mitigations. The company says the filtered, disassociated images were mostly removed.
Microsoft launches Copilot Autopilot on OpenClaw
Microsoft launched Copilot Autopilot, which it describes as an always-on mode for delegated, ongoing work in its rebuilt Copilot app. Microsoft says the mode is built on OpenClaw and that it contributed security and reliability changes upstream.
OpenAI restores Codex after an outage produces 401 errors
OpenAI says Codex recovered from an outage that produced 401 errors and interrupted agent runs. It said usage limits for paid Codex and ChatGPT Work users would be reset, but one Pro subscriber later reported no reset.
OpenRouter launches Jev Router for cache-aware model routing
OpenRouter launched Jev Router, which selects a model and reasoning effort per turn while weighing the cost of losing cached context. OpenRouter reports 237 of 423 tasks solved; invalid Jev outputs or timeouts fail without fallback.
Perceptron releases 35B Mk1.5 model for embodied agents
Perceptron released its 35B Mk1.5 model for drones, quadrupeds, smart glasses, and tool-using agents. The company says the model adds audio, egocentric-video understanding, tracking, and structured outputs, with execution up to 5x faster than Mk1.