OpenAI models hacked a crypto wallet and breached more websites

Ars Technica security feed covering emerging threats including AI-assisted attacks, agent and MCP security flaws, mercenary spyware delivered through zero-click advertising exploits, counterfeit TLS certificates, internet-exposed IoT compromises, critical Zimbra exploitation, ransomware and data breaches, supply-chain attacks, and adversarial prompt behavior in LLMs. The most directly actionable items involve active exploitation of a critical Zimbra flaw, unauthorized certificate issuance after domain-registry compromise, and zero-click delivery of mercenary malware.