Microsoft launches ASSERT for AI behavior testing
Microsoft has released ASSERT, an open-source framework that appears to help developers evaluate AI system behavior for specific product and service requirements.
Sections
AI & Models
529 storiesMartin Scorsese joins AI startup Black Forest Labs as adviser
Martin Scorsese has joined AI startup Black Forest Labs as an adviser, though the director is reportedly using the technology solely for storyboarding purposes.
Microsoft launches Scout, an AI assistant for Microsoft 365
Microsoft is launching Scout, an AI assistant for the Microsoft 365 ecosystem built on the OpenClaw framework that requires a GitHub Copilot subscription.
Microsoft releases open source standard for AI agent control
Microsoft has released the Agent Control Specification (ACS), an open-source standard designed to help developers define and enforce governance policies for AI agents across various frameworks.
OpenAI expands Codex tools to target enterprise knowledge workers
OpenAI is expanding its Codex toolset with new plug-ins and features, targeting enterprise knowledge workers as usage grows more than three times faster than developer adoption.
Anthropic scales Project Glasswing to 15+ countries
Anthropic is expanding its Project Glasswing AI security initiative to about 150 organizations across more than 15 countries to help secure critical infrastructure codebases.
WindBorne Systems launches AI model to improve weather forecasting
WindBorne Systems launched WeatherMesh-6, an AI weather model that claims to outperform leading government systems by integrating proprietary balloon sensor data directly into its predictions.
Box founder Aaron Levie warns of AI psychosis among tech CEOs
Box founder Aaron Levie suggests tech CEOs are prone to "AI psychosis," while users turn to alternatives like DuckDuckGo amid frustration with Google's AI search integration.
Developers are increasingly refusing to work without AI
Developers increasingly refuse to work without AI, but research suggests this reliance may create long-term maintenance costs and fail to improve overall code quality.
A guide to the language of artificial intelligence
Artificial intelligence is changing the world, and this living glossary helps professionals navigate the evolving language and terminology used by industry leaders.
Cognition raises $1 billion at $26 billion valuation
Cognition raised $1 billion at a $26 billion valuation, with CEO Scott Wu emphasizing that its Devin AI agent should augment, rather than replace, human software engineers.
Anthropic releases Opus 4.8 with new dynamic workflow tool
Anthropic has released Opus 4.8, its newest AI model, just 41 days after its predecessor, amid competitive pressure from OpenAI and Google.
Apple prepares AI-powered Siri overhaul for iOS 27
Apple could introduce a new Siri app in iOS 27, leveraging Google’s Gemini AI and local models to bring generative AI to its 2.5 billion device install base.
Why recursive AI remains a distant goal for the industry
Startups are chasing recursive self-improvement as a goal, though experts warn the industry is not very close to achieving truly autonomous, self-upgrading AI systems.
Why Google’s AI models fail at simple word-level tasks
Google’s AI Overviews are struggling with basic spelling and counting tasks, a limitation stemming from the token-based architecture of large language models.
ElevenLabs launches Music v2 with mid-track genre switching
ElevenLabs launched Music v2, a model capable of mid-track genre switching, nearly 10 months after its initial music-generation release.
Human Archive raises $8.2M to train robots with gig worker data
Human Archive raised $8.2 million to collect egocentric video data from gig workers for physical AI training, despite facing regulatory scrutiny and pushback from local service platforms.
Google’s new AI agents are locked behind an only $100 monthly plan
Google has unveiled a new ecosystem of AI agents, including Gemini Spark, but is currently restricting access to its only $100-per-month Ultra plan subscribers.
OpenAI model disproves 80-year-old geometry conjecture
OpenAI’s latest reasoning model has disproved an 80-year-old geometry conjecture, marking a significant step forward in AI's ability to handle complex, multi-step mathematical reasoning.
Tesla’s FSD software gains second European approval
Tesla has launched its FSD (Supervised) software in Lithuania, marking its second European market as the company seeks to scale its driver-assistance technology.