•Pronto, a Bengaluru-based home services startup, has launched a pilot programme where cleaners wear head-mounted cameras during jobs for a small subset (0.1%) of customers.
•The initiative aims to address customer anxiety about unfamiliar workers in their homes but has raised serious privacy concerns due to potential AI training use.
•Videos are anonymised, no audio is recorded, and footage is deleted within 48 hours, but “derived datasets” like body movement tracking are retained for AI training.
•Rival platforms Urban Company and Snabbit have distanced themselves from such practices, emphasising their commitment to customer privacy.
•The controversy spotlights the growing demand for “egocentric” data in AI training, with some labs reportedly paying $4–$10 per hour for such recordings.
•Legal experts note that while anonymised data may not fall under India’s DPDP Act, the country lacks a clear governance framework for non-personal data.
•Public backlash on social media reflects deep unease over balancing AI innovation with personal privacy in home environments.
BENGALURU: A Bengaluru-based on-demand home services startup, Pronto, has sparked widespread concern after confirming that a small subset of its customers can opt into a programme where household cleaners wear visible, head-mounted cameras during jobs. The initiative, aimed at addressing customer anxiety about unfamiliar workers entering their homes, has triggered a privacy debate after it was revealed that the recorded footage could potentially be used to train artificial intelligence (AI) and robotics systems.
Pronto cofounder and CEO Anjali Sardana stated that the pilot programme covers only 0.1% of users and is designed for those who feel uneasy about leaving their homes unattended. “They worry about what's happening in their home during their booking. Something may be stolen or broken, or work may not be done properly,” she explained. The company claims that videos captured are anonymised, no audio is recorded, and all footage is deleted within 48 hours. However, Pronto confirmed that “derived datasets” from these recordings—including “key point mapping” data tracking body joints and hand movements—are retained for AI training purposes.
Customers who choose this option are charged an additional ₹29 per booking, and workers are also compensated extra for participating. Despite these measures, the company’s disclosures have done little to ease public concern. Social media platform X has seen posts criticising the move, with one user writing, “recording inside your house to train AI! This is scary,” while another commented, “Trust is the cornerstone of any consumer/service business and Pronto just lost it.”
The controversy has prompted rival platforms to distance themselves from such practices. Urban Company cofounder Abhiraj Bhal emphasised that his company prioritises customer privacy and does not engage in recording activities for AI training. Similarly, Snabbit founder Ayush Agarwal confirmed that his company is not involved in any such initiatives.
Behind the debate lies the rapidly growing market for “egocentric” or first-person data, which is essential for training physical AI systems. Unlike text-based AI models, physical AI requires exposure to real-world environments such as kitchens, utensils, shelves, and human movements. Some AI labs globally are reportedly paying between $4 and $10 per hour for such data, according to a founder in the field.
The legal landscape surrounding this issue remains unclear. Nikhil Narendra, a partner at Trilegal, noted that anonymised household data might fall outside the scope of India’s Digital Personal Data Protection Act (DPDP Act). However, he cautioned that India still lacks a comprehensive governance framework for non-personal data, leaving significant ambiguity about data usage and privacy protections.
As the conversation unfolds, the debate highlights the tension between technological advancement and individual privacy, raising questions about how long the “front door” will remain the last bastion of personal privacy in an increasingly digitised world.
