IBM says attacks that started with exploitation of public-facing applications rose 44% year over year in its 2026 X-Force Threat Index. The report also says vulnerability exploitation made up 40% of incidents observed in 2025 and that more than 300,000 ChatGPT credentials were exposed by infostealers.
Anthropic published a Mar 6, 2026 policy for vulnerabilities identified with Claude. The framework sets a 90-day default disclosure window, a 7-day target for actively exploited critical bugs, and human review requirements before reports go out.
The Anthropic-Mozilla collaboration that spread on Hacker News disclosed that Claude Opus 4.6 found 22 Firefox vulnerabilities, 14 of them high-severity. The durable lesson is not autonomous magic but faster defender workflows built around validation, triage, and reproducible evidence.
Microsoft’s Security Dashboard for AI entered public preview on February 13, 2026. The dashboard aggregates Defender, Entra, and Purview signals to give security leaders a unified view of risk across AI apps, agents, models, and MCP servers.
Microsoft Threat Intelligence said on March 6, 2026 that attackers are now using AI throughout the cyberattack lifecycle, from research and phishing to malware debugging and post-compromise triage. The report argues that AI is not yet running fully autonomous intrusions at scale, but it is already improving attacker speed, scale, and persistence.
Anthropic published a March 6, 2026 case study showing how Claude Opus 4.6 authored a working test exploit for Firefox vulnerability CVE-2026-2796. The company presents the result as an early warning about advancing model cyber capabilities, not as proof of reliable real-world offensive automation.
AWS introduced Amazon Connect Health on March 5, 2026 as a healthcare-specific agentic AI product for administrative and point-of-care workflows. Patient verification and ambient documentation are available now, while appointment management, patient insights, and medical coding are in preview.
A LocalLLaMA post details recurring Whisper hallucinations during silence and proposes a layered mitigation stack including Silero VAD gating, prompt-history reset, and exact-string blocking.
Anthropic says a March 4 Department of War letter designates it as a supply chain risk, but argues the scope is narrow and will challenge the action in court.
On January 13, 2026, Anthropic announced an expanded Labs organization focused on experimental Claude products. The company is formalizing a two-track model: fast frontier experimentation and separate operational scaling for reliable customer-facing products.
Anthropic published a March 5, 2026 research report introducing an observed-exposure metric that combines theoretical AI task feasibility with real Claude usage, finding mixed early labor-market signals.
At Mobile World Congress on February 28, 2026, NVIDIA and major global telecom and infrastructure partners announced a joint commitment to open and secure AI-native 6G platforms. The initiative ties operator adoption, ecosystem standards, and AI-RAN execution into a single coalition roadmap.