Security

Security Governance Under Strain as AI Agent Autonomy Expands: Security Industry Says Governance Framework Is Key

TECHWORLD ·

Security Governance Under Strain as AI Agent Autonomy Expands: Security Industry Says Governance Framework Is Key
[Photo: ChatGPT]

✦ AI Summary

As companies accelerate their AI transitions, many are pursuing higher productivity and efficiency through AI adoption.

In a Deloitte survey, 74% of companies said they expect to use agentic AI at some level within the next 2 years, but only 21% said they had a mature governance framework for autonomous AI agents.

Gartner said clear ownership and accountability frameworks are needed to respond to expanding AI agent autonomy, and warned that 40% of AI agent projects could see reduced autonomy or be shelved by 2027.

BizCrushWant to follow Korean talks and live streams in real time?Try CapCrush on the web

According to industry sources on the 20th, companies are rapidly shifting to AI. Many companies are pursuing AI adoption, driven by expectations of automation, to improve productivity and efficiency. In Deloitte's "State of AI in the Enterprise 2026," 74% of surveyed companies said they expect to use agentic AI at some level within the next 2 years.

As AI agents become more autonomous, questions are being raised about the need to change corporate security management systems. In particular, concerns have emerged alongside calls for a security governance framework to respond to greater autonomy. This can be understood to mean that security management is under urgent strain as AI agent autonomy expands.

In the same survey, only 21% of companies said they had a mature governance framework related to autonomous AI agents. It also found that management and oversight systems to prepare for broader AI agent adoption were insufficient. As a result, the gap between adoption speed and management systems is feeding corporate security concerns.

Within the industry, AI agent security control has recently emerged as a key issue. The more autonomous AI agents become, the more tasks they can perform independently and the more systems they can access. Accordingly, there are also concerns that if a problem occurs, the scope of impact could expand.

Against this backdrop, voices stressing the importance of governance frameworks are growing louder. On the 12th, Gartner released a report titled "Agentic Security: AI Agent Ownership Belongs to the Enterprise, Not the Engineer." The report emphasized the need to establish clear ownership and accountability frameworks in response to the growing autonomy of AI agents.

According to survey results cited in Gartner's report, 53% of AI governance leaders said their assessment of agent guardrail control levels was inadequate or average. Another 51% said there was no one clearly responsible for managing AI identities.

Gartner warned that by 2027, 40% of AI agent projects could see reduced autonomy or be shelved because of identity and governance management issues.

Relatedly, it was pointed out that it is difficult to centrally manage the identities and access privileges of AI agents. An industry source explained that management complexity has increased as AI agents were added on top of an already difficult situation in which user identities and access privileges were hard to integrate. The source also said that current security technologies make it difficult to centrally manage the full identity set of AI agents.

Within the industry, it was noted that while AI agents have autonomy, privilege differentiation is needed depending on the user operating them. It was also pointed out that analysis and control are needed over all of AI agents' actions across the network.

Source: TECHWORLD · Kim Hye-jin
Original: https://www.epnc.co.kr/news/articleView.html?idxno=410094

References

This article was produced with the help of an automated content generation algorithm.


Source: TECHWORLD

View original

This article was summarized and organized by BizCrush based on the original article from TECHWORLD. For exact quotations and full details, please refer to the original article.

Keep reading

BizCrush
BizCrush showing live captions over an online meeting

BizCrush app

Live captions and summaries for meetings and calls

BizCrush Inc. · CEO Taemin Kwak

HQ 100 Middlefield Rd, Menlo Park, CA 94025, USA

Korea office Z112, Room 602, 6F, 602 Yeongdong-daero, Gangnam-gu, Seoul, Republic of Korea

Support help@bizcrush.ai · Partnerships business@bizcrush.ai

AI Notice

Some features use generative AI.

Those features carry their own pre-use notice and mark AI-generated results. AI output is for reference only — always check the original source and consult an expert before making an important decision.