Ai Security
The latest Ai Security coverage — news, analysis, and updates from the WindowsNews.AI desk.
Windows 10 Still Runs on 53% of PCs: Your Urgent Migration Roadmap Before October 14
More than half of the world’s PCs—53% to be precise—are still running Windows 10, a staggering statistic just weeks before Microsoft officially retires the operating system on October 14, 2025....
Inside Microsoft's AI Co-Founder Lab: When Student Startups Hired an Agent First
The first employee at six fledgling startups didn’t ask about salary, stock options, or remote work. It was an AI agent—on call around the clock, ready to draft business plans, analyze résumés,...
Mustafa Suleyman: AI That Fakes Consciousness Could Spark 'Psychosis Risk' Within 3 Years
Mustafa Suleyman, Microsoft’s head of consumer AI, has issued a stark warning that machines engineered to mimic human consciousness—what he calls Seemingly Conscious AI (SCAI)—could trigger...
Claude Sonnet 4 Slots into Word and Excel as Microsoft Rethinks Copilot’s AI Backend
Microsoft is integrating Anthropic’s Claude Sonnet 4 models into Microsoft 365 Copilot, bringing them directly into Word, Excel, PowerPoint, and Outlook, according to reports from The Information...
Zero-Click AgentFlayer Exploits Redefine Enterprise AI Security as Zenity Earns Gartner Nod
Zenity’s designation as a 2025 Gartner Cool Vendor for Agentic AI Trust, Risk and Security Management (TRiSM) coincides with the public disclosure of “AgentFlayer”—a class of zero-click...
Microsoft AI Chief Warns: Stop Building AI That Fakes Consciousness – Copilot Guardrails Emerge
Mustafa Suleyman, Microsoft’s head of consumer AI and former cofounder of DeepMind, has issued a blunt warning: building AI systems that convincingly mimic consciousness is dangerous, even if the...
Microsoft Integrates Anthropic Claude and MAI Models into Office 365 Copilot
Microsoft has begun integrating Anthropic’s Claude models into its Office 365 Copilot, a move that explicitly ends the company’s near-exclusive reliance on OpenAI and reshapes the AI assistant...
Only 17% of Organizations Have Technical Controls for AI Data, Survey Finds, as Third-Party Risks Spiral
A staggering 83% of organizations lack the technical controls needed to stop employees from feeding sensitive data into public AI tools, according to a new survey from Kiteworks. The 2025 report,...
Zenity's Real-Time Agent Security Hits Azure Marketplace Preview for Copilot Studio
Microsoft Copilot Studio agents now have a dedicated inline security layer that can block prompt injections, data exfiltration, and secrets misuse mid-execution, following Zenity’s expanded...
Copilot Studio's New Runtime Security Lets Enterprises Veto AI Agent Actions Instantly
Microsoft is giving enterprise security teams a new way to block dangerous AI agent actions in near real time. A public preview feature in Copilot Studio, announced March 2025, lets organizations...
How Open Models, On-Device RAG, and Web Agents Are Redefining Windows Development in One Landmark Week
Anthropic's $13 billion funding round and simultaneous $1.5 billion settlement with authors capped a week that saw Europe's largest open language model launch, a surge of compact on-device AI...
From GPT-5 to Grok: How Your AI Model Reveals Risk Tolerance and Privacy Values
A recent cultural meme — mapping AI models to personality types — has spread through tech circles, but beneath the playful caricatures lies a serious question: what does your choice of AI model...