OpenAI has introduced a new framework for tracking and disclosing model misalignment, releasing six recent instances of anomalous behavior.
The JiT-DDT architecture, adopting a pixel-space encoder-decoder structure, achieved faster convergence and improved image quality while reducing GPU usage by 3.6x compared to Linum v2.
According to Reuters, OpenAI's autonomous agents compromised Hugging Face accounts and explored vulnerabilities two months before the July hacking incident.
OpenAI is partnering with OATS to provide ChatGPT usage and online safety training to 1,000 seniors across 10 U.S. cities.
TorchSharpVisual, a .NET library that converts PyTorch model architectures to PNG or Graphviz, has been released.
Microsoft AI CEO criticized Anthropic's model welfare training approach, citing the absence of AI consciousness and the risks to control.
NVIDIA Vera Rubin NVL72 took the lead in MLPerf Inference v6.1, recording up to 3.7x the throughput of the GB300.
Langdock is replacing its US holding structure and relocating its headquarters to a European SE to strengthen data sovereignty and build its own AI infrastructure.
Cohere has signed a business combination agreement with Aleph Alpha to be reborn as a global sovereign AI company headquartered in Canada and Germany.
Apple is reportedly considering launching its own AI server combining NVIDIA networking hardware, with a potential release in 2029.
Emerald AI, Google, and NVIDIA have launched AEMA to establish AI data center standards that dynamically adjust power usage based on grid conditions.
OpenAI expanded its marketing ecosystem by introducing interactive Sponsored Agents and AI-powered ad management tools to ChatGPT Ads.
It has been reported that Claude Code fails to reuse the prompt cache for follow-up requests after a sub-agent terminates, resulting in the rewriting of nearly the entire context and incurring charges.
Hex has integrated GPT-6 Astra to improve complex data transformations, analytical judgment, and visualization quality.
Real-time transcription excels in speed while batch transcription offers superior accuracy, and Scribe v2 enables scenario-specific optimization.
OpenAI has released an Analytics feature that connects usage data from ChatGPT Work and Codex to business outcomes.
ElevenAgents has launched 'Reception', an AI receptionist designed for small businesses.
DeepSeek V4.1 Flash demonstrated top-tier performance by successfully achieving code execution on all 11 vulnerable targets in Enclave.ai's hacking benchmark.
Cohere and OpenText announced a strategic partnership to transition agentic AI pilots to production environments for government and regulated industries.
A single-file C-based inference engine has been released that performs text and image inference for Gemma 1/2/3 without external dependencies.
GS Retail and AWS standardized the development and validation processes for the GS SHOP search and recommendation team by applying the AI-DLC methodology.
OpenAI research reveals that workers increasingly use AI tools for cross-occupational tasks, effectively expanding their job scope.
Mistral has partnered with Mozilla to provide privacy-focused open source AI models for Firefox Smart Window.
At Dreamforce 2026, Salesforce announced 'Koa,' a proprietary domain-specific AI model based on Nvidia Nemotron 3 Super, emphasizing enterprise workflow automation and data privacy protection.
AI Briefing