# 92% of Indian firms hit by AI agent security incidents in 2025

2026-11-18T11:42:51+00:00 | Technology | Indian Opinion Desk

Corroboration: 2 independent outlets
Outlets counted: thehindubusinessline.com, timesofindia.indiatimes.com

A ServiceNow report has found that 92% of Indian organisations surveyed experienced a security incident linked to AI agents in the year to November 2026, with only 36% able to contain an incident within 24 hours. The survey of 500 security leaders, titled ‘Deployed but Ungoverned: The State of Agentic Security in India’, reveals that just 18% of firms have unified visibility of their AI agents from a single interface, and only 19% have one named person accountable for governing agentic AI security. The report also found that 37% of organisations discovered AI agents running in production that were never registered with their security teams. While 51% have automated real-time detection of unusual agent behaviour, 67% cannot automatically respond when an agent behaves anomalously. The findings follow a CERT-In advisory on AI agents issued in October 2026 and come as existing rules require Indian organisations to report cyber incidents within six hours of detection.

## Indian Opinion Analysis

The three outlets report ServiceNow's findings with near-identical figures and attributions, all leading on the 92% incident rate. NDTV and The Times of India highlight the 36% containment rate in their opening lines, while Business Line emphasises the gap between confidence and control, noting that half of leaders call their security 'established' yet only 19% have a single accountable person. The Times of India adds the report's title and a direct quote from ServiceNow's managing director, which the other two omit. Business Line provides the most detail on 'shadow AI' and the CERT-In advisory from October 2026, absent elsewhere. No outlet adopts a clear slant, the coverage is uniform straight reporting of a vendor survey. The differences are only in what supplementary figures each chooses to include. The CERT-In advisory, mentioned only by Business Line, is the nearest concrete next step on government oversight.

## Coverage

Coverage: 3 sources, 3 neutral
- timesofindia.indiatimes.com (neutral report) <https://timesofindia.indiatimes.com/india/9-in-10-indian-firms-hit-by-ai-agents-in-past-year-only-a-third-contained-it-within-a-day-survey/articleshow/135961029.cms>
  Opens with 92% and 36%, includes quote from ServiceNow MD and report title
- thehindubusinessline.com (neutral report) <https://www.thehindubusinessline.com/companies/indian-firms-hit-hard-by-ai-agent-security-gaps-struggle-to-respond-in-time/article71715454.ece>
  Opens with 92% and 36%, adds CERT-In advisory context and shadow AI detail

Background, not counted: ndtv.com <https://www.ndtvprofit.com/technology/92-of-indian-companies-faced-ai-agent-security-incident-last-year-most-couldn-t-stop-it-in-a-day-12343950#publisher=newsstand>

This brief was synthesised by AI from the 3 sources linked above, so one read covers every framing they carry.

Tags: CERT-In, Kulmeet Bawa, ServiceNow
Canonical: https://indianopinion.org/92-of-indian-firms-hit-by-ai-agent-security-incidents-in-2025/
License: Summary and commentary (c) Indian Opinion, reusable with attribution. Facts belong to the linked sources.
Cite: https://indianopinion.org/92-of-indian-firms-hit-by-ai-agent-security-incidents-in-2025/#story-in-brief
Gate decision: pass (pass: linked qualifying outlets 2 of 3, relevance 8)
Freshness: pass (fresh: newest qualifying source 2026-11-18 11:30Z (feed), 0.2 h old)
Image: photograph. Source: https://static.toiimg.com/thumb/msid-135961173,imgsize-68248,width-400,height-225,resizemode-4/in-a-recent-survey-an-alarming-92-of-indian-organizations-reported-security-incidents-involving-ai-agents.jpg.

How this brief was made: an AI model read the reports linked above and wrote this summary and analysis, which were published automatically. Published briefs are sampled every hour by an automated quality check; the editor verifies its findings and approves corrections, and corrected briefs carry a dated correction line. Stance labels are editorial classifications of how each outlet framed this story, assigned by the same model, not ratings of the outlets. We do no original reporting. Methodology: https://indianopinion.org/ai-use-policy/
