AI Agents Shift Oversight to Identities and Vendor Dependency
What This Is About
Gartner organises the ten trends into three thematic areas: new AI models beyond text processing, control of autonomous AI agents, and building long-term resilient IT architectures. According to Gartner, the central challenge is that AI systems increasingly operate outside established control structures. They generate software faster than humans can review it, create new security risks and increase dependency on individual technology providers.
Why This Matters for the Board and Executive Management
Two of the ten trends directly affect overall management and overall supervision under Art. 716a OR. First, managing identities and access rights for autonomous agents: Gartner expects that by 2028, one quarter of security breaches in companies will result from the misuse of AI agents. Second, strategic dependency on individual providers: Gartner predicts that by 2030, more than 75 per cent of companies outside the United States will pursue digital sovereignty strategies. These focus particularly on reducing geopolitical risks and avoiding vendor lock-in. Both are forecasts from a market research firm that itself offers consulting on these technologies. They do not replace your own risk assessment, but they show where oversight questions are shifting.
Questions for Executive Management
- Does the company manage and monitor identities and access rights of AI agents separately from human user accounts, and how is misuse of such agent access detected?
- Does the company have a strategy to reduce dependency on individual AI and cloud providers, for example through options to switch critical software, hardware or models?
Sources
- Gartner nennt KI-Governance und digitale Souveränität als zentrale Technologietrends (Netzwoche, 04.11.2026)