AIがシステムを操作する時代、SaaSはどう変わる？ freeeが示すMCPの課題と可能性
2026-04-09
ITmedia
The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Gartner predicts that by 2028, 25% of enterprise generative AI applications will experience at least five minor security incidents annually, up from 9% in 2025, due to increased use of agent-based AI and Model Context Protocol (MCP). Risks include information leaks and inadequate security controls, prompting calls for stricter safeguards.[AI generated]
Why's our monitor labelling this an incident or hazard?
The event involves AI systems (generative AI applications) and discusses risks related to their use and integration with external components via protocols like MCP. However, the harms described are potential and forecasted rather than actualized. No specific AI Incident has occurred yet; instead, the article highlights plausible future security risks and advises on preventive measures. Therefore, this qualifies as an AI Hazard, as it concerns credible risks that could plausibly lead to AI Incidents in the future if not properly managed.[AI generated]