⚠️ DISCLAIMER: Ministry of Cyber Affairs is an independent news, guides and directory service on cybercrime and cybersecurity, for victims and investigators. It is NOT an official government or ministry portal.
Website Scam CheckerNew
Back to News

Suggestions from China's Deepseek for securing India's Online Exams

8 June 2026
Source:Deep Seek Prompt on securing CBT
Share
Suggestions from China's Deepseek for securing India's Online Exams

"Kendriya Pariksha Kendras” & "National Secure Exam Platform (NSEP)" are the most significant suggestions

New Delhi

A simple prompt on Deepseek, regarding how to secure online exams in India, returned a comprehensive 10-point technical and institutional blueprint built after extensive review of global practices, explicitly including China’s Gaokao infrastructure.

The plan centers on the National Secure Exam Platform (NSEP), codenamed “Suraksha Exam OS”, a read-only, digitally signed, kiosk-mode operating system that locks down commodity hardware.

USB ports, external storage, screen sharing, and background apps are disabled at the kernel level.

The OS verifies its own integrity at boot and cannot be altered by exam centers.

Key pillars

  • Offline-first architecture with encrypted exam packages synced locally 30 minutes before the test via MPLS, 4G/5G, or satellite, and responses uploaded only after the session ends. This eliminates the live-internet attack surface that has plagued past breaches.
  • Continuous Aadhaar-linked biometric chain, registration, entry turnstiles, passive webcam-based face authentication during the exam (flagging gaze aversion, extra faces, or objects near the face), and final re-verification at submission.
  • Dynamic question generation from a high-security National Item Bank using a psychometric AI engine. Every candidate receives a unique but statistically equated form with shuffled sequences and options, rendering mass memorization or copying ineffective. Threshold cryptography ensures no single party can decrypt papers early.
  • Blockchain-based immutable audit trail for every critical event, biometric matches, question access, proctor interventions, creating forensically admissible records.
  • 3,000 permanent “Kendriya Pariksha Kendras” (starting with 500 upgrades) featuring thin-client terminals, multi-angle night-vision cameras, redundant networks, RF shielding or approved jammers, and a 200-meter no-device perimeter with drone surveillance.
  • National Command Centre co-located with CERT-In and the Ministry of Home Affairs, aggregating edge AI behavioral alerts in real time.
  • Central Exam Integrity Authority (CEIA), an autonomous body modeled on the Election Commission with standard-setting powers, unannounced audits, blacklists, and investigation/prosecution authority.
  • Stringent vendor controls, source-code escrow, mandatory red-teaming, and the long-pending Public Examination (Prevention of Unfair Means) Bill with 5–10 year sentences for organized fraud.

The 36-month phased plan with Aadhaar biometrics and AI proctoring mandates, followed by pilots of the secure OS and full scaling of permanent centers is the low handing recommendations.

More in Cybersecurity

Related articles

Cybersecurity

OpenAI Disclosed ChatGPT User Content, Including Prompts and Files, in 88 Government Requests

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.

12 Nov 2026
Cybersecurity

CyberStrike AI - Detection for Defenders: Scripting, Tempo and Residue

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.

4 Nov 2026
Cybersecurity

Cybersecurity Checklist in the Agentic Hacking Era

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.

2 Nov 2026