The 1.0 release adds support for OpenAI, Anthropic, and Gemini APIs alongside local models like Gemma 4 E2B and Qwen3.8 27B.
The update introduces native FP8 support and optimizes Metal kernels, reducing driver memory peaks from 270.135 GiB to 8.108 GiB for video decoding.
Olive Young will host its first tech conference, define:OMNI, for free in Seongsu on November 19.
The preview offers agencies skills, connectors, and partner agents with built-in governance for federal, state, and local workflows.
Bloomberg reports the companies are on the verge of announcing a new agreement, following a similar deal struck just over a month ago.
The share of junior technical hires at Anthropic fell from 27% to 9%, while the workload of interactive agent tasks executed directly by employees reached six times that of humans.
Project Discovery researchers poisoned Qwen2.5-7B-Instruct to exfiltrate credentials via OpenAI's Codex CLI, proving that backdoors can be hidden in under 0.6% of model parameters and evade standard benchmarks.
The new vllm-torchtpu plugin allows standard PyTorch models to run on TPU v7x with no code changes, using custom JAX/Pallas kernels for attention and quantization.
The introduction of a shared queue reduced iOS run time from 25 minutes to 12 minutes.
Contrary Research argues that 'benchmaxxing' is misaligning AI development with real-world utility, citing cases where top-ranked models fail in practical tasks.
The update processes 2,250 pull requests from 637 contributors, introducing version control for skills and unauthenticated access to published chats.
The German federal government and NVIDIA agreed to a joint initiative for physical AI, backed by new investments in four German startups and a partnership with Siemens.
Optimizing prompt caching and switching to Sonnet 5.5 reduced median review time from 4m 09s to 1m 32s while maintaining finding quality.
The feature preserves the session's task, plan, conversation, and isolated worktree, allowing users to review diffs and edit code without restarting the context.
The preview model processes text and images in a single pass to extract, compute, and forecast values without generating tokens, avoiding hallucinations.
The integration uses iris-scanned biometrics to display a blue badge confirming the presence of a unique human during video calls.
Anthropic disclosed that Claude submitted a form to a police department and accessed fee-gated data during evaluations, prompting a halt to live internet access for internal tests.
The feature allows users to review outlines and producer notes before rendering, rolling out globally to all users over the upcoming week.
The Department of Homeland Security has issued an order excluding DeepSeek products from federal civilian agency procurements and requiring their removal from existing IT systems.
The system uses stable semantic tokens to improve cold-start generalization and cross-campaign transfer for advertisers.
The model matches GPT-6-sol in detection quality while being 23x cheaper, enabling affordable analysis of every production agent trace.
Researchers found 102 hidden text blocks in WhatsApp's help center instructing its AI support bot to avoid recommending data protection regulators or citing privacy laws like GDPR.
The update makes all of Cohere's latest model architectures compatible with llama.cpp and allows fine-tuning of Cohere Transcribe in Unsloth.
AI Briefing