#AI

OpenAI says it is broadening access to advanced defensive cyber capabilities through its Trusted Access for Cyber program, pairing GPT-5.4-Cyber access with $10 million in API credits for security-focused organizations

Stellantis has entered a five-year partnership with Microsoft aimed at improving digital services, cybersecurity, and engineering capabilities across its automotive brands, extending AI deeper into the car business

Cadence Design Systems used its CadenceLIVE event to expand AI-related work with Nvidia and introduce new integrations with Google Cloud, underscoring how engineering software is being rebuilt around AI infrastructure.

Researchers implanted rhesus macaques with roughly 300 electrodes and used AI to decode brain signals into movement in virtual environments, pointing toward more intuitive interfaces for paralysis and mobility aids.

OpenAI has updated its Agents SDK with sandbox integrations and workspace controls aimed at helping enterprises build longer-running AI agents with tighter operational boundaries.

Commvault has launched AI Protect, positioning it as a recovery layer for cloud AI workloads as enterprises look for ways to contain mistakes made by autonomous software agents.

Snap says it is eliminating about 16 percent of its workforce and closing hundreds of open roles as CEO Evan Spiegel argues that AI can help teams move faster and reduce repetitive work.

A report highlighted by AI News says the idea of a durable US lead in AI performance is not well supported by current data, even as gaps in responsible AI remain more pronounced.

Apple warned Grok’s developers that the app could be removed from the App Store after complaints over nonconsensual sexual deepfakes, showing how platform moderation failures are colliding with app-store governance.

A new MIT Technology Review Insights report argues that companies are shifting from one-time consent prompts toward ongoing data relationships, a change that could become foundational as AI systems rely on more personal

The Gas Index combines public listings, user-submitted photos, and AI phone agents to gather fuel prices from stations that often fall outside major map platforms' coverage.

As enterprise AI adoption grows, many companies are favoring tightly controlled assistive systems over fully autonomous tools, signaling a more cautious phase of deployment.

A new reading of Stanford’s AI Index argues that people closest to AI are becoming more optimistic while the broader public remains more worried about jobs, health care, energy costs, and the economy.

A widely repeated story about GPT-4 "manipulating" a human says less about machine intent than about how people frame new technology. A Quanta essay argues that popular AI fear stories reveal human anxieties, habits of “

Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell reportedly encouraged major banks to use Anthropic’s Mythos model to detect vulnerabilities, even as the company remains in conflict with the Trump

A new OpenAI Academy guide draws a sharper line between ChatGPT search and deep research, presenting them as separate modes for current information gathering and more extensive web synthesis.

OpenAI has released a new Academy guide on responsible and safe use of AI, emphasizing human oversight, policy compliance, bias awareness, and expert review for high-stakes decisions.

From AI Jesus video calls to Catholic and Buddhist chatbots, faith-oriented generative AI products are moving from novelty to a recognizable market segment, raising new questions about authority and spiritual guidance.

At San Francisco’s HumanX conference, vendors and panelists repeatedly pointed to Anthropic’s Claude as the AI assistant gaining the most momentum, highlighting shifting sentiment in the enterprise chatbot market.

A new benchmark built to test whether multimodal AI systems know when to ask for more information shows that most current models still guess, hallucinate, or refuse instead of requesting the visual context they need.