AI governance, in practice.
Notes on keeping personal data out of large language models — what the DPDP Act actually asks of you, how prompts leak, and where a control point belongs.
- 6 min read
Shadow AI: Why Blocking It Fails and What to Do Instead
Shadow AI is the unapproved use of AI tools at work. Why bans push it further out of view, what it actually costs you in visibility and evidence, and how to replace prohibition with a governed path.
Shadow AIGovernancePolicy - 8 min read
DPDP Act Compliance for AI: What Indian Companies Need to Know
DPDP Act compliance for AI tools: what the Digital Personal Data Protection Act, 2023 requires of Indian companies whose teams send personal data to LLMs, and the controls that support those obligations.
DPDP ActComplianceAI Governance - 7 min read
How to Prevent PII Leaks When Your Team Uses ChatGPT and Other LLMs
How to prevent PII leaks to LLMs: why prompts carry personal data out of your network, why blocking and training fail, and how detection and tokenisation at the egress point stop the leak without stopping the work.
PIIData ProtectionLLM Security - 7 min read
Why On-Premise AI Gateways Matter for Data Privacy
An on-premise AI gateway inspects and masks prompts before they cross your network boundary. Why that placement decides whether a privacy control actually prevents disclosure or merely relocates it.
On-PremiseArchitectureData Privacy