Ai Security
The latest Ai Security coverage — news, analysis, and updates from the WindowsNews.AI desk.
ASCII Smuggling in Gemini AI Sparks Security Debate Over Prompt Injection
Google's controversial decision not to patch a newly discovered "ASCII smuggling" vulnerability in its Gemini AI has ignited a fierce debate about how to properly secure generative AI models...
AI Hallucination Drops 50% in 2025: Enterprise Governance Evolves
The year 2025 marks a significant turning point in the evolution of artificial intelligence, with major improvements in reducing AI hallucinations across leading language models. While the...
250 Malicious Documents Can Backdoor LLMs, Evading Standard AI Safety Checks
A groundbreaking study from Anthropic reveals that as few as 250 malicious documents can implant reliable backdoor behaviors in large language models, challenging fundamental assumptions about AI...
Clipboard Data Leaks: How Employees Accidentally Expose Sensitive Data Through AI Tools
A disturbing new security trend is emerging in workplaces worldwide: employees are inadvertently leaking sensitive corporate data through generative AI tools by simply copying and pasting...
Microsoft Copilot Expands with Harvard Health Content for Trusted Medical Answers
Microsoft has significantly enhanced its AI assistant's healthcare capabilities through a strategic licensing agreement with Harvard Medical School's consumer arm, Harvard Health Publishing. This...
Agentic AI faces new threats; Sentinel Copilot and Foundry debut at Ignite 2025.
Microsoft's security strategy for 2025 represents a fundamental shift in how organizations will protect their AI ecosystems, with Agentic AI moving from experimental technology to critical...
Microsoft Copilot Integrates Harvard Health Content for Trusted Medical AI
Microsoft is taking a significant step toward making AI-powered health information more reliable by integrating curated content from Harvard Health Publishing directly into Copilot. This strategic...
Microsoft Licenses Harvard Health Content for Copilot: A Strategic Shift in AI Safety and Model Diversification
Microsoft's reported licensing agreement with Harvard Medical School's Harvard Health Publishing represents a pivotal moment in the evolution of AI assistants, marking a strategic shift toward...
Microsoft 365 AI Automation: How Copilot Transforms Everyday Productivity
Microsoft's consumer-focused initiative to "Unlock productivity with AI automation" represents a fundamental shift in how artificial intelligence integrates into daily workflows, positioning Copilot...
Harvard Health's vetted medical library now powers Microsoft Copilot's health responses
Microsoft's groundbreaking partnership with Harvard Health Publishing represents a significant milestone in the evolution of AI-powered health information, bringing authoritative medical content...
OpenAI Sora Dominates iOS Charts Amid AI Safety and Copyright Debates
OpenAI's groundbreaking video generation tool Sora has taken the mobile world by storm, exploding onto the iOS charts with an estimated 627,000 downloads across the U.S. and Canada during its first...
Your Employees Are Pasting Corporate Secrets into ChatGPT—And There's No Audit Trail
Nearly half of all employees are now using generative AI at work, but their method of sharing data with chatbots is flying under the radar of security teams. A new report from browser security firm...