Cybersecurity

Security research, breaches, privacy, vulnerabilities, crypto and security risks, and trust.

Cybersecurity14 days agoBy Jey Geethan

Anthropic launches opt-in OSS Scanner with unreviewed vulnerability reports

Participating open-source projects can receive free, periodic security scans, but Anthropic says model-generated findings reach maintainers without human review. Enrollment requires a buildable project, while Anthropic says the service is intended for teams able to triage reports.

Read article

A circular archive holds photographs, papers, film and a disc beneath a broad terracotta cover with a green pull tab.

Cybersecurity21 days agoBy megan

Apple plans extra controls for macOS Full Disk Access

Apple wants a more explicit user action before apps receive broad access to Mac data. It has not announced when or how the new controls will work.

Read article

Cybersecurity29 days agoBy megan

Google Says PageBreak Found More Than 500 XSS Flaws Across Its Web Apps

Google says its internal PageBreak security agent found more than 500 cross-site scripting flaws by pairing AI-led investigation with deterministic exploit validation. The project also shows why secure-by-design frameworks still matter.

Read article

Cybersecurityabout 2 months ago

AIR Raises $50M to Continuously Vet AI-Agent Add-Ons

AIR has emerged from stealth with $50 million across two rounds to sell continuous vetting and policy enforcement for AI-agent skills, plugins and MCP servers. Its security and adoption figures remain company claims, while public deployment and accuracy data is limited.

Read article

Cybersecurity3 months ago

Apple’s 26.6 Security Updates Span iPhone, iPad, Mac and More

Apple released a broad set of 26.6 security updates on July 27, 2026, alongside fixes for older macOS branches. The advisories justify prompt action, but IT teams should still test representative devices and avoid implying that every listed flaw is actively exploited.

Read article

Cybersecurity3 months ago

Open Secure AI Alliance Targets Security for Models and Agents

Nvidia and dozens of partners have launched the Open Secure AI Alliance to share security research and tools spanning agent harnesses, identity, model formats and vulnerability scanning.

Read article

Cybersecurity4 months ago

GitHub Turns Private Vulnerability Advisories Into Dependabot Fixes

GitHub's innersource security advisories let enterprises privately warn internal repositories about vulnerable components and trigger Dependabot fixes.

Read article

Cybersecurity5 months ago

AI Concerns for Software Security: Why Faster Code Is Creating Faster Risk

AI coding tools are speeding development, but new reports warn of insecure code, prompt injection, shadow AI and faster vulnerability exploitation.

Read article