96% of IT Professionals Adopt AI, With 59% Planning Agentic Deployments and Security Oversight Gaps
What Happened — A global survey of 700 data analysts and 700 IT leaders (Alteryx, May 2026) found that 96 % now use AI in their daily work. While only 49 % are frequent users, 59 % expect to deploy AI agents within the next 12 months and more than half are willing to grant those agents unrestricted data access. 44 % say human oversight is essential, yet the security implications of such access were not addressed in the study.
Why It Matters for TPRM —
- Unrestricted AI‑agent access creates a new attack surface for data leakage and insider‑type misuse.
- Validation of AI outputs is emerging as a critical skill; vendors lacking this capability may deliver inaccurate or risky decisions.
- Organizations must embed AI‑specific controls into third‑party risk programs before agents become production‑grade.
Who Is Affected — Enterprises across all sectors that rely on external SaaS, cloud‑hosted analytics platforms, or MSP‑delivered AI services.
Recommended Actions —
- Inventory any third‑party AI agents or “agentic” services in use.
- Require contractual clauses for data‑access limits, audit logs, and human‑in‑the‑loop oversight.
- Verify that vendors have documented AI‑output validation processes and incident‑response plans for AI‑related misuse.
Technical Notes — No specific vulnerability or exploit was disclosed. The risk stems from unrestricted data access by AI agents (potential misuse, inadvertent data exposure, or model‑driven exfiltration). Organizations should treat AI agents as a third‑party dependency and apply the same security hygiene as for any external software component. Source: ZDNet Security – 96% of IT pros use AI now