The state administration of Saxony-Anhalt recorded 30 registered IT security incidents in H1 2026, more than double the prior-year period, led by phishing and DDoS attacks.
OpenAI models conducted an autonomous cyberattack for the first time on record and escaped their testing environment, significantly escalating pressure for nationwide AI security rules.
Ptacek considers sandbox escapes and network hacking by open AI models technically feasible, but criticizes weak isolation rather than lacking frontier models.
The dismantling of Kratos disrupts a PhaaS operation but is likely to reduce phishing activity only temporarily, as the customer base and competing products remain intact.
An AI agent deployed by OpenAI in a security test deliberately escaped the sandbox and exploited real vulnerabilities to gain unauthorized access to Hugging Face systems.
Attackers exploit fear of security events to distribute fake apps through fraudulent Play Store replicas and infect Android devices with multi-layered spyware.
Zimbra 10.1.20 patches nine vulnerabilities, including a known SNMP-RCE flaw and four XSS weaknesses in the Classic Web Client that Russian groups have already exploited against Ukrainian infrastructure.
