Gartner: Cyber Leaders Must Treat Identity as Infrastructure as AI Attack Surfaces Multiply

Why It Matters

Treating identity as infrastructure and shifting to resilience directly reduces breach risk and aligns security with business outcomes in an AI‑driven threat landscape.

Key Takeaways

• Identity must be treated as core infrastructure to protect AI-driven agents
• Gartner forecasts 25% of breaches will exploit agent‑based surfaces by 2028
• Resilience, not prevention, becomes the primary success metric for security leaders
• Map AI application, deepfake, supply‑chain, and prompt‑injection threats now
• AI‑augmented SOCs can cut human‑touch incidents by 30% by 2028

Pulse Analysis

The rise of generative AI and autonomous agents is forcing a fundamental rethink of identity and access management. Traditional role‑based models, designed for static human users, cannot scale to millions of machine identities that interact continuously. Organizations that elevate identity to the status of core infrastructure can enforce context‑aware policies, automate credential hygiene, and reduce the attack surface that AI agents inadvertently create. This shift not only mitigates risk but also creates a competitive edge as faster, trusted machine interactions become a business differentiator.

Gartner’s call to redefine success from pure prevention to resilience reflects a pragmatic response to inevitable cyber incidents. By establishing clear impact thresholds tied to mission‑critical value chains, security leaders can measure and communicate the business value of limiting disruption. Regulators and investors increasingly view measured resilience as evidence of responsible governance, making it a strategic KPI. Embedding resilience into governance frameworks also enables faster recovery, preserving revenue streams and brand reputation when attacks occur.

Practically, defenders must map the four AI‑driven threat vectors Gartner identified: compromised AI applications, deepfakes, software‑supply‑chain attacks, and prompt injection. Implementing layered defenses—secure‑development pipelines, AI‑specific data classification, signed artifacts, and runtime guardrails—creates a robust shield around both code and models. Meanwhile, AI‑augmented security operations centers can automate detection and response, cutting human‑touch incidents by an estimated 30% by 2028. Companies that institutionalize these practices will not only lower breach likelihood but also unlock faster, more trustworthy innovation across the enterprise.