Cybersecurity, vulnerabilities, threats, and defense
OpenAI’s new Lockdown Mode for ChatGPT blocks data exfiltration from prompt injection attacks by restricting outbound network requests. It’s available for all personal and Business accounts, but comes with significant feature tradeoffs.
DSIT’s Nick Woodcraft revealed at Infosecurity Europe 2026 how the UK government manages vulnerabilities across 500,000+ domains — by focusing on outcomes over technical jargon and drip-feeding fixes instead of overwhelming teams.
Attackers exploited a flaw in Meta’s AI-powered High Touch Support tool to hijack over 20,000 Instagram accounts, bypassing two-factor authentication entirely. The campaign ran undetected from April 17 to May 31, 2026.
Cybercriminals have registered nearly 19,000 FIFA-themed domains since January, targeting World Cup fans with phishing, ticket scams, travel fraud, and malware. The FBI and Meta have both issued warnings ahead of the 2026 tournament.
Over 2.5 million student loan borrowers had their personal data — including Social Security numbers — exposed through a breach at Nelnet Servicing. The incident comes at the worst possible timing, coinciding with federal loan forgiveness announcements.
Cisco disclosed its seventh SD-WAN zero-day of 2026 — CVE-2026-20245 — a command injection flaw allowing root access on Catalyst SD-WAN Manager. No patch is available yet, and there’s no workaround.
Nelnet Servicing exposed 2.5M student loan records including SSNs in a breach that went undetected for weeks. Here’s what borrowers need to do now.
A new 269-page House bill would regulate frontier AI with independent audits, open-source security grants, and a $300M NIST center. But the fight over state preemption could tear it apart.
Cisco disclosed its 7th actively exploited SD-WAN zero-day in 2026. CVE-2026-20245 allows root command execution with no patch available yet. Here’s what defenders need to know.
A bipartisan House bill proposes sweeping AI regulation including $300M for NIST’s CAISI, open-source security grants through CISA, and mandated AI-assisted vulnerability review — but faces fierce opposition over language that would preempt state AI laws.
