AI and privacy: An analysis finds that the Digital Personal Data Protection (DPDP) Act, 2023 does not address AI training on personal data.

Telecom grievances: TRAI issued the Telecom Consumers Complaint Redressal (Fourth Amendment) Regulations, 2026 on 2 November 2026.

• Public data left out: The Act does not apply to publicly available personal data (Section 3(c)(ii)), including data scraped from social media and inferred data such as emotions and behaviour.
• No sensitive-data category: Passwords and health data are treated like a name or a shopping interest, a step back from the IT Act, 2000 and its 2011 Rules.
• Silence on AI: AI models can memorise and regurgitate training data, so “anonymised” data can be used to re-identify and profile people.
• No right to object: Indian users cannot stop their public data from being used to train AI, while Meta gives EU users an opt-out.
• Weak erasure: The right to be forgotten covers only data given with consent, and is exempt for research, archiving and statistics (Section 17(2)(b)). It is unclear if AI training counts as research.
• Timeline: The Rules were notified in November 2025, with the main provisions in force from May 2027.

• Who is covered: Internet Service Providers with more than 10,000 subscribers.
• Easier complaints: Consumers can complain or appeal through the provider’s mobile app, besides the existing channels.
• 24×7 help: Complaint Centres must work round the clock. AI agents can be used, but human help must be available at least 16 hours a day.
• Accessibility: Special helplines and accessible apps for persons with disabilities.
• Faster appeals: The time to appeal falls from 30 to 15 days, and the Appellate Authority must decide within 15 days.
• Accountability: Providers must run a feedback survey after each complaint, report to their CEO or Board every quarter, and set up a Consumer Grievance Redressal Oversight Committee.
• Enforcement: TRAI can audit providers and impose financial penalties for non-compliance.

• Digital protection: Both stories ask how users are protected when services run on data and automation.
• Privacy: A fundamental right under Article 21 needs rules that cover AI-era data use.
• Trust: Fast, accessible and accountable grievance systems build trust in digital services.

• Balance: Too tight a rule can slow AI, and too loose a rule weakens privacy.
• Enforcement: DPDP provisions are phased in, and the Data Protection Board is new.
• AI agents: Human oversight must be real, not only on paper.