Pronto's camera pilot sparks India privacy debate

Pronto, a Bengaluru-based home-services startup, was reported to have used video footage recorded inside customers' homes to train physical AI and robotics systems, according to Business Standard and the OECD.ai incident monitor. Journalist Harsh Upadhyay alleged on X that some service professionals used "small outward-facing cameras during select opt-in jobs," as summarized in reporting by Business Today and other outlets.

Reporting by Business Today quotes a Pronto post on X: "Unless you have opted-in and paid for the program personally, the Pro doesn't come to the house with a camera. Opt in is not one time, it has to be affirmed before each booking. By default there is no camera involved, and when there is, it's impossible to miss." Business Today also reports the company said the pilot covered "0.1% of customers" and that it had spent "months to ensure full DPDP compliance."

Business Standard notes Pronto handles over 25,000 household service orders daily, providing scale to any data-collection pilot. The OECD.ai monitor frames the episode as an AI hazard, saying the practice raises plausible privacy and human rights concerns though no direct harm has been reported. Multiple media outlets including The Hindu Business Line and Hindustan Times covered the controversy and reported rivals such as Urban Company and Snabbit distancing themselves from similar practices. Business Today also reports Pronto raised $20 million in a recent extension round led by Lachy Groom.

Companies collecting in-home audiovisual data for training face repeatable technical challenges that matter for practitioners. Consent must be collected, stored, and auditable on a per-session basis; implementing per-booking opt-in workflows increases product complexity. Data-minimization and deidentification for video are technically hard: simple cropping or blurring often degrades training signal, while advanced anonymization (pose-only representations, synthetic replacements) requires additional pipeline steps and validation. Secure storage, retention policies, and access controls are necessary when footage may contain sensitive cues about household composition, children, or medical devices.

Industry reporting and expert commentary emphasize a broader legal and regulatory gap in India. The Hindu Business Line and other outlets cite experts warning that India's evolving data protection framework leaves ambiguities around the legality and acceptable uses of in-home recordings for AI training. The OECD.ai monitor places the episode in a category of AI hazards that attract rights-based scrutiny even absent a proven incident. Observers following the sector are likely to view this as part of a global trend where convenience-driven data-collection pilots run ahead of clear regulatory interpretations.

For practitioners and compliance teams, track formal guidance or enforcement from Indian regulators on the Digital Personal Data Protection regime and any sector-specific clarification about biometric, audiovisual, or in-home data. Watch for regulator complaints, class actions, or consumer-safety notices that could define permissible practices. Also monitor competitor policies and procurement language from large platforms and clients, since reputational pressure and customer terms can shape operational constraints before formal law changes. Finally, technical teams should evaluate privacy-preserving training alternatives (edge-only inference, synthetic data, pose-based representations) and consent-auditing tooling as part of risk mitigation.