03-Oct-2026 Microsoft, GitLab and OpenAI Face Escalating AI Security Threats

Hacked dAily is the first AI-driven cybersecurity podcast, created by Cytadel Cyber and published daily. Each episode delivers concise, decision-ready security news for CISOs, security leaders, and executives. Today’s episode covers: 1. A critical remote code execution flaw in middleware used across financial and government environments could allow attackers to take control of trusted systems. Rapid patching and enhanced monitoring are essential to prevent disruption, data exposure, and wider network compromise. 2. Microsoft’s official X account was briefly hijacked to promote a Clippy-themed cryptocurrency. The incident demonstrates how compromised corporate profiles can amplify scams, damage trust, and mislead large audiences. 3. GitLab patched a critical vulnerability affecting self-hosted AI features that could enable malicious activity in organizations’ development environments. Teams should apply updates immediately and review exposed AI configurations to protect code, data, and developer accounts. 4. Microsoft researchers warn that attacks against AI systems can now be developed in minutes, shrinking the window for defenders to respond. The finding reinforces that AI is an immediate business risk, with potential consequences including model manipulation and sensitive data exposure. 5. Researchers identified a Mythic command-and-control technique that uses a legitimate cloud file service to exchange encrypted data. The approach could make malicious activity harder to detect, highlighting the need for stronger monitoring of cloud services and data movement. Hacked Daily is sponsored by Cytadel, an offensive cyber assurance company specialising in AI-powered attacks. Cytadel’s expert-led Red Team Assessments follow real attack chains across people, identity and technology to verify whether they can become AI fraud or ransomware disruption. We verify your defences before attackers do. Learn more at cytadel.co.uk.