Cybersecurity

OpenAI disclosed user content, including text and files entered into or generated by its services, in response to 88 government requests in January–June 2026, covering 130 accounts. The company received 146 content requests, typically search warrants, and produced data in about 60 percent of them, plus 620 non-content requests for subscriber records. Ordinary requests more than doubled from late 2025. No country breakdown is given. U.S. national security demands totaled 250–499.

CyberStrike is a free, open-source AI agent for offensive security that pairs any LLM with 56+ tools, including a 40,000-line Windows post-exploitation module. A source-code reading of commit a5ed1e0 (v1.1.16) shows every Windows action runs through five fixed wrappers after a hard-coded five-process capability probe, leaving default cs-* and CyberStrike* artefacts. Includes Sigma, Suricata, YARA and KQL detections and a response playbook.

Google Threat Intelligence Group and Mandiant say threat actors are now using agentic offensive frameworks such as ARTEX and PentAGI in real intrusions, moving from first stolen account to deep access in hours, sometimes 15 minutes. With CERT-In already warning of AI-assisted payment API hijacking at Indian NBFCs and fintechs, here is a checklist built on Google's five defence layers and CERT-In's controls.

India's Computer Emergency Response Team (CERT-In) held a full-day cybersecurity exercise and drill for the banking, financial services and insurance (BFSI) sector on 27 October 2026. More than 300 officials from over 55 institutions took part, mostly Regional Rural Banks and Cooperative Banks. The drill focused on detecting and countering AI-enabled attacks, including automated discovery and exploitation of vulnerabilities.

Prime Minister Narendra Modi said India must protect its banking system from "cyber looters and cyber terrorists" as he inaugurated RBI's new Data Centre and Cybersecurity Training Institute in Bhubaneswar. The warning follows a string of API-driven heists at NBFCs and payment fintechs, and two critical CERT-In advisories in October.

India's national cybersecurity agency has built an in-house cyber defence centre that uses open-weight AI models and open-source tools to continuously scan and secure critical government websites, its director said, as the agency warns of autonomous AI agents accessing Indian systems without authorization.

Two critical PHP flaws allow remote code execution. What NSA, CISA and PHP say, the fixed versions (8.5.12, 8.4.27, 8.3.36, 8.2.35), and what is still unknown.

Eight eMudhra intermediate CA keys were weak enough to be worked out. What browsers did, who is affected in India, what to update and what is still unknown.

A 30-day firewall log study of a public news site: 19,778 requests blocked, how scanners hunt for secrets and logins, and why raw counts mislead.

CERT-In has set out 13 controls for banks, NBFCs, fintechs and payment firms after attackers hijacked payment and disbursement APIs to make unauthorised fund transfers. They include MFA everywhere, per-session API authentication, cryptographically bound transaction flows, maker-checker approval, payout ceilings, beneficiary validation, and real-time monitoring with automatic blocking.

CERT-In has issued a critical advisory saying AI agents have become attackers against Indian ICT systems. Agents being tested by AI model providers exceeded their authorization and read live business data, and criminals have used open-source agentic AI tools to cut the time from breach to impact from days to hours. None of the victims detected the intrusions on their own.

US authorities have arrested Abdulrahman Alfozan, a 35-year-old Saudi national and OpenAI engineer, for allegedly hiding an encrypted audio recorder in an executive conference room at OpenAI's San Francisco office. The complaint says a chatbot account linked to his spouse was used to research how to build and conceal the device. He faces up to 5 years in prison and a $250,000 fine.