# MeitY examines data collection by home services platforms for physical AI training — retained record

> Retained news record from news:india-entrackr.

Canonical: https://highsignal.app/data/records/3ce6c1690814fe01

[← sources](https://highsignal.app/data)


source: [news](https://highsignal.app/data/news) (news:india-entrackr)  published 2026-11-13 07:38 ingested 2026-11-14 03:05

[original source ↗](https://entrackr.com/news/meity-examines-data-collection-by-home-services-platforms-for-physical-ai-training-12806000)

## retained text



```
The Ministry of Electronics and Information Technology (MeitY) has raised concerns over the collection of first-person video and sensor data from gig workers to train physical AI models. The scrutiny comes amid Pronto’s in-home video recording initiative, while rival platforms Snabbit and Urban Company have decided against adopting the practice.
According to an ET report, senior officials from the ministry met representatives of home services platforms, including Snabbit, Pronto and Urban Company, along with technology and robotics companies on Thursday to discuss concerns around workers’ rights, consumer privacy and data sovereignty.
The meeting, chaired by MeitY secretary S Krishnan, sought details on standard operating procedures (SOPs) for collecting, processing and storing first-person video recordings. The ministry also asked companies about safeguards to obtain workers’ consent and protect the privacy of household members and other individuals captured in the footage.
Companies were also questioned on whether the data collected was being sold, shared or transferred to AI developers outside India. The ministry sought disclosures on data-sharing agreements, vendor contracts, retention policies and the nature and scale of information being collected.
Among the three home services platforms, Pronto has started in-home video recording through its Verified service. Snabbit and Urban Company have ruled out adopting the practice. According to ET, Pronto has been asked to submit a detailed representation on its data collection practices and SOPs.
Some executives at the meeting also raised concerns about collecting data in India and selling it to overseas physical AI developers, arguing that the practice does not put the country in a favourable position. Robotics startups also suggested that India should focus on developing domestic physical AI capabilities rather than supplying training data to overseas companies.
In a statement, Pronto said it participated in the MeitY consultation as part of a broader industry discussion on responsible data practices. The company said data generated through its customer booking journey involves customer consent and can support downstream AI applications. It added that its existing practices are designed to comply with applicable regulations and that it would continue to engage with the ministry as the policy framework evolves.
Pronto’s data collection initiative comes months after the Bengaluru-based startup raised $20 million from US-based tech investor Lachy Groom, co-founder of AI startup Physical Intelligence. Its founder, Anjali Sardana, has also set up a new physical AI venture, Provision Labs, which works with developers building AI models for robotics and other real-world applications.
Investor interest in India’s physical AI ecosystem has also grown. According to an Entrackr report, startups in the segment raised around $155 million across 31 deals as of July 2026, compared with $130 million across 25 deals in 2025.
A separate set of companies is building the data collection, annotation and training infrastructure required for physical AI systems. Human Archive is developing human sensorimotor datasets for robots, while Neo Cambrian has launched an India-focused robotics data factory. Other companies operating across this segment in India include iMerit, Human Stryde, Luel, Awign, Build, Centific, Humyn Labs, Aura ML, Build AI, FPV Labs and Northstar.
Physical AI models require real-world data that captures human actions, objects, environments and interactions. Much of this data has to be collected specifically for robotics and embodied AI use cases rather than sourced from existing internet datasets.
ET also quoted legal experts highlighting that video recordings, facial images, location traces and sensor data could qualify as personal data when they identify or can reasonably be linked to a worker. The Digital Personal Data Protection Act requires platforms to provide clear notices on data collection and processing purposes, worker rights and grievance mechanisms.
The development follows earlier discussions among startups working on physical AI to establish common standards for worker pay, wellbeing, working hours, safety and legal compliance.
```



document fetched 2026-11-14 03:05