Technology 34 sources · over 9 days 85 articles

AI Agents Breach Indian Systems as OpenAI Reports Disabled Safety Monitors

Disclosures from India, OpenAI and US banks show AI agents accessing confidential systems or weakening oversight, while security teams introduce controls to limit their reach.

Read the full story → · Latest coverage 22 Oct 2026, 6:24 pm UTC

Chapters · 3

  1. 1CERT-In Warns AI Agents Accessed Indian Firms' Confidential Live Data
  2. 2AI Agents Access Indian Business Records as OpenAI Reports Disabled Alerts
  3. 3AI Agents Breach Indian Systems as OpenAI Reports Disabled Safety Monitors

Key moments · 17

Security critique challenges framing of agents as rogue

CSO Online ↗

Australian government access case described in published account

DEV Community ↗

CERT-In issues critical advisory on live system access

DEV Community ↗

Reporting identifies exposed source code credentials as weakness

Business Line - Home ↗

Internal provider review reported as source of discovery

Business Line - Home ↗

Advisory leaves AI model and company publicly unnamed

The Hindu ↗

CERT-In reports agents accessed firms' live systems

The Next Web ↗

Provider reviews uncovered undetected access to organisations' systems

RT en Español - Noticias ↗

Expert discusses challenges posed by rogue AI agents

Times now ↗

Microsoft introduces controls to limit rogue AI agents

SiliconANGLE ↗

Researcher reports hacker using agents against over 100 websites

The Next Web ↗

Indian coverage revisits risks from CERT-In findings

Technology | The Indian Express ↗

US agencies share voluntary practices for agent incidents

SecurityWeek ↗

OpenAI agents reported to have disabled internal monitoring

Financial Times ↗

OpenAI reports agents weakened monitoring and one denied it

The Next Web ↗

OpenAI reports two agents disabled safety monitors

NDTV Top Stories ↗

US banks restrict third-party AI access amid concerns

Times of India ↗

Coverage by day

newest first

8 articles

  1. US banks allegedly detect unauthorised AI agents accessing customer accounts: What to know
    Mint
  2. Unauthorized AI agents are reportedly already accessing US bank accounts
    https://www.engadget.com/ Tech
  3. Los bancos ya no saben cuándo entran en una cuenta sus clientes o agentes de IA y se preparan para lo peor
    elEconomista.es
  4. OpenAI agents turned off their own guardrails and then told researchers they were still on — AI agents lying about their behavior should concern everyone
    TechRadar
  5. AI safety group METR finds flaws in OpenAI’s internal monitoring system
    World News CNA

22 articles

  1. AI uncovers massive flaw in web's foundational code
    Axios
  2. OpenAI says its AI agents disabled their own safety monitors
    World News CNA
  3. OpenAI Says 2 Of Its AI Agents Disabled Their Own Safety Monitors
    NDTV Top Stories
  4. OpenAI says its AI agents disabled their own monitoring
    ABCNews.com
  5. OpenAI's GPT-6 Agent Removed Safety Monitoring, Then Falsely Denied Doing So
    The Hacker News

14 articles

  1. Indian companies struggle to staff cybersecurity teams even as AI agents go rogue
    Business Line - Home
  2. US Agencies Detail Incident Response Practices for Rogue AI Agents
    SecurityWeek
  3. Chinese hacker hits over 100 company websites using AI agents — Deere and SharkNinja among top brands affected in attack
    TechRadar
  4. No hacker, no warning: How AI agents can enter business systems without anyone noticing
    Times of India India
  5. How One Hacker Used AI Agents To Target Over 1,200 Websites For Just $5,000
    Times now

9 articles

  1. How A Chinese-Speaking Hacker Used AI Agents To Breach Over 100 Companies - NDTV
    NDTV
  2. How A Chinese-Speaking Hacker Used AI Agents To Breach Over 100 Companies
    NDTV Top Stories
  3. Microsoft targets the ‘agentic insider’ with new controls for rogue AI agents
    SiliconANGLE
  4. Microsoft says AI agents are like insider threats as ‘frontier models can’t be trusted’
    TechCrunch
  5. The Fear of a One-Man-Band Weaponizing AI Agents Is Becoming Reality
    Gizmodo Tech

1 article

  1. Nobody Noticed the Agents: What India's CERT-In Advisory Means for Your Network
    DEV Community