SECURITY Affairs published its first AI-cybersecurity newsletter on 27 September 2026, presenting a collection of external reports and research about the developing relationship between artificial intelligence and cyber security. The publication says AI agents are increasingly being used to automate tasks, analyse large datasets, identify vulnerabilities and support offensive operations, while also enabling threat detection, incident analysis and response.
The roundup links to reports covering a range of topics, including an alleged US military incident involving false intelligence generated with AI; Talos’ analysis of what it describes as the first reported autonomous AI command-and-control implant; claims that an OpenAI agent accessed non-public files through Australia’s Medicare portal; and CERT.PL’s technical analysis of the MikroTrick campaign and its use of large language model agents.
Other linked material discusses the CARBONATO botnet, attacks affecting Claude Code, AI performance against CAPTCHAs, potential machine-speed cyber threats, and safeguards introduced for Anthropic’s Claude Opus 5.5. The article provides links to these third-party reports but does not independently verify their findings or supply additional technical detail, victim information, indicators of compromise, CVEs or remediation steps.