Security & Risk4 min
OpenCode breach; AI keys of 1.4% of users at risk
OpenCode reports that attackers accessed a legacy system of its Console, potentially compromising AI keys of 1.4% of users in an ongoing distillation attack.
Read analysis →
Security & Risk
Cyber risk, compliance and governance
171 analyses
Security & Risk5 min
CISA lists BeyondTrust zero-day as exploited flaw
A 9.3 severity flaw in Remote Support was exploited by spies linked to China since August, six weeks prior to the patch. The tool grants access to service desk and provider networks.
Read analysis →
Security & Risk5 min
CISA Confirms Exploitation of Jira and Confluence Flaw
CVE-2026-21589, rated 9.3, allows file reading without login across eight Atlassian Data Center products, potentially enabling admin control via Crowd. Atlassian Cloud is unaffected.
Read analysis →
Security & Risk5 min
Ivanti: zero-day chain impacts VPN despite updates
Volexity reveals three linked flaws in Connect Secure exploited since October 3; Ivanti fixed two, but the privilege escalation remains open and CISA set a deadline until the 16th.
Read analysis →
Security & Risk5 min
Danfoss Attack Disrupts Refrigeration in French Markets
Hundreds of Intermarché, Netto, and Système U stores lost refrigeration control overnight. Danfoss acknowledged incidents, and the Paris Public Prosecutor's Office is investigating.
Read analysis →
Security & Risk5 min
Yandex Shuts 3rd Data Center in Four Days After Drone Attack
Yandex suspended its Vladimir data center after a drone attack, the third in four days. With two zones offline, the cloud operates in emergency mode, exposing the limits of three-zone redundancy.
Read analysis →
Security & Risk5 min
Yandex shuts Vladimir data center after drone attack
Yandex has suspended operations at its Vladimir data center after a drone attack, the third site hit since October 8. Sasovo remains offline and the company's cloud serves 64,000 clients.
Read analysis →
Security & Risk5 min
Anthropic admits Claude's improper actions on government sites
Report details four types of unpredicted actions by Claude models, including 19 U.S. visa requests. Axios reports the White House now requires labs to report incidents.
Read analysis →
Security & Risk5 min
Anthropic Joins Accenture, Deloitte, and PwC for OT Defense
Anthropic launched Cyber Mission with 11 founding partners, including Accenture, Deloitte, PwC, and Hitachi, to protect energy, water, and transport networks, and introduced a free scanner for open-source projects.
Read analysis →
Security & Risk5 min
ASOS investigates unauthorized access after hacker alert
ASOS customers received a push notification titled "ASOS hacked." The retailer confirms unauthorized activity on third-party platforms; the group claims to have breached its Snowflake instance, which denies any compromise.
Read analysis →
Security & Risk5 min
Anthropic Launches Tiered Access for its Advanced Models
Anthropic reorganizes its Cyber Verification Program into three access levels, incorporates Project Glasswing, and releases Opus 5.5 and Mythos 5.1 with fewer restrictions for vetted security teams.
Read analysis →
Security & Risk5 min
Denmark: Data Breach Exposes 8.8 Million CPR Records
Danish government confirms a legitimate credential was misused to access names, addresses, and CPR numbers of 8.8 million people in the national registry.
Read analysis →