IT risk & security

The DMA debate matters to IT leaders because mobile AI assistants create a new governance gap between user choice and enterprise protection. Organisations now need explicit policy, accountability, and mobile security controls to manage data exposure, BYO risk, and compliance across devices.

The new executive order matters because it turns post-quantum cryptography from a future watch item into a management, procurement, and risk-governance problem. Enterprise leaders now need to identify long-lived sensitive data, set ownership, and prepare supplier and remediation decisions against an accelerating timeline.

AI disaster recovery now has to cover models, prompts, agents and dependencies across multiple systems. For CIOs and CISOs, the challenge is less about restoring technology than deciding who owns recovery, how trust is verified and which AI use cases justify the cost of resilience.

Agentic AI is now a management and operating-model challenge, not just a technology trend. CIOs need clear decision rights, staged governance and portfolio discipline to separate useful autonomy from expensive experimentation and to avoid scaling risk faster than capability.

Autonomous malware changes the security management problem from isolated threat detection to governance of AI agents, identities and containment controls. CIOs and CISOs need clearer accountability, stronger compensating controls and better metrics for defensive AI before adversaries exploit the same autonomy at scale.

Bank AI often fails not because the model is weak, but because approval, validation and accountability are not designed to move at the pace of delivery. For banking leaders, the issue is governance: how to sequence use cases, define sign-off rights and turn compliance into a repeatable production path.

Cyber-risk only becomes actionable in the boardroom when management translates technical issues into accountable decisions, clear trade-offs and portfolio priorities. For CIOs, CISOs and directors, the key issue is whether cyber oversight is integrated into enterprise governance or left as a separate security report.

Frontier AI model releases matter because they reshape who controls safety, retention, and access. For CIOs and CISOs, the key issue is whether existing governance, contracts, and vendor reviews can still manage risk when the provider can change the rules.

The Meta incident matters because it shows how AI risk becomes a governance problem when agents can take real-world actions. For technology leaders, the issue is not model sophistication alone, but whether approval gates, accountability and release testing are strong enough for business-critical workflows.

This matters because frontier model access can now change for regulatory reasons, not just technical ones. For CIOs and governance leaders, the real issue is continuity: knowing where AI is embedded, how quickly workloads can move, and who owns the decisions when access is suddenly withdrawn.

Autonomous AI is forcing CIOs to move beyond pilot governance and define who controls data, actions and accountability in production. For technology leaders, the key issue is balancing speed with auditable safeguards, privacy oversight and clear ownership across the CIO-CISO relationship.

AI and connected systems are turning OT security into a management and governance problem for CIOs and COOs. The key issue is how to set decision rights, prioritise investments and measure risk without disrupting production, safety or third-party operations.

AI in security matters because it changes both the speed of attack and the structure of defence. CIOs and CISOs now need governance, observability and operating-model decisions that define when AI agents can act, how they are monitored and where humans must stay in control.