1. SEBI–RBI push “Demat 2.0”: tokenised corporate bonds, CBDC settlement and AI supervision
SEBI and RBI have moved ahead with Demat 2.0, a pilot that uses distributed-ledger technology for corporate bonds, CBDC-based settlement and smart contracts for asset servicing. SEBI Chairman Tuhin Kanta Pandey said the first phase involving three issuers has been completed and the next phase will extend to secondary-market trading. SEBI is also working to implement IOSCO’s supervisory toolkit for AI in the securities market.
Why it matters: This brings together DLT, tokenisation, CBDC, smart contracts and AI governance inside regulated financial-market infrastructure. For BFSI CISOs, auditors and fraud-risk teams, the security focus now expands to smart-contract assurance, wallet/key management, transaction integrity, cyber-resilience and auditability of AI-supported market systems.
2. UP Police pushes AI from reactive policing toward “anticipatory policing”
At the 3rd Police Technology Summit 2026 in Lucknow, UP Police Additional Director General (Technical Services) Bhagwan Swarup Srivastava argued that AI can help policing move beyond reactive and even predictive approaches toward an anticipatory model, where threats are identified before escalation. The discussion covered AI analytics, cyber forensics, digital-evidence management, drones and predictive technologies.
Why it matters: The real policing opportunity lies in fusing CCTV/video analytics, CDR/IPDR, cybercrime complaints, financial trails, OSINT and crime records. But anticipatory policing also requires governance: explainability, bias testing, human review, lawful data use and evidentiary integrity must accompany deployment.
3. More than 60 women allegedly recruited into mule-account network through fake loan promises
Pimpri-Chinchwad Police have arrested two men after investigators uncovered an alleged scheme in which more than 60 women in Goa were induced to provide bank-account access after being promised loans. Police say the accounts were subsequently supplied to an international cyber-fraud network and used to route fraud proceeds. The investigation began after a victim lost ₹83.80 lakh in an alleged foreign-IPO investment scam.
Investigators reportedly traced substantial flows through the accounts and are examining links with cybercrime cases in several states.
Why it matters: Mule-account procurement is becoming a distinct organised-crime layer. Banks and investigators increasingly need to detect sudden behavioural changes, newly activated accounts, rapid inward/outward transfers, common devices/IPs, beneficiary clustering and accounts controlled by third parties. Financial-trail analytics can expose the network rather than just the first-layer account holder.
Proposal for Conducting Cyber Crisis Drill, Tabletop Exercise (TTEx) & CCMP Readiness Exercise
4. Delhi under major anti-drone and multi-agency security shield for BRICS Summit
With the 18th BRICS Summit beginning today in New Delhi, Delhi Police has deployed a large security apparatus reportedly involving around 25,000 personnel, more than 100 NSG commandos, anti-drone measures and secured routes/hotels. Separately, Delhi Police imposed a blanket prohibition on drones and numerous other aerial platforms across the capital from 10–14 September, using powers under Section 163 of the BNSS.
Why it matters: VIP and critical-infrastructure security is becoming a counter-UAS intelligence problem, not merely a physical-perimeter problem. Effective capability increasingly requires RF detection, radar/electro-optical correlation, drone classification, operator geolocation, lawful mitigation and subsequent forensic examination of seized drones/controllers.
5. Anthropic reports state actors and cybercriminals using AI for espionage, malware and weapons-related activity
In the major global cybersecurity development, Anthropic’s latest threat-intelligence report says it disrupted a range of malicious AI-use campaigns involving cyberattacks, espionage, surveillance, malware development and weapons-related research. Anthropic linked one campaign to Russia-associated Midnight Blizzard and described large-scale attempts by Chinese AI firms to extract or replicate Claude capabilities.
The findings show that sophisticated actors are increasingly using frontier models not merely for drafting phishing emails but for malware modification, reconnaissance, analysis and operational automation.
Why it matters: AI-enabled cybercrime is moving toward an agentic threat model. SOCs and government cyber units should plan for machine-speed reconnaissance and attack adaptation, requiring behavioural detection, stronger identity controls, sandboxing of AI agents, human authorisation for consequential actions and improved forensic logging.
Today’s signal
Three shifts are becoming especially clear: policing is becoming AI-assisted and anticipatory; financial crime prevention is moving toward network-level analytics rather than individual-account investigation; and AI itself is increasingly becoming both a security tool and an attack platform.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
