Eyewear startup Lenskart has begun piloting AI-enabled self-eye tests at select stores, its founder and CEO Peyush Bansal said during the company’s Q1 FY27 earnings call, raising privacy and data protection concerns. Can companies use eye test data to train their proprietary AI models?
“This quarter, our AI-enabled self-eye test entered pilot stores, built on one of India’s largest recorded eye test data sets,” Bansal said.
Lenskart already provides AI-enabled remote eye testing. In a letter to shareholders, Bansal said remote testing stores have now scaled to 786 and home tests are live across 1,200+ pincodes.
Responding to an analyst’s question about the accuracy of these remote eye tests, Bansal said results are far more consistent than those of physical eye tests. The company leverages Tango Eye, the AI-based computer vision startup it acquired in 2023, to monitor remote tests using CCTV cameras.
Lenskart is feeding users’ biometric data to AI: Lenskart collects and stores facial recognition data of millions of Indian users. Since January, the company has been asking customers, whether visiting its stores or logging on to its app, to complete facial scans. In Q3 FY26, the company said it captured 100,000 face scans daily. This figure may have gone up now.
During the Q1 FY27 earnings call, Bansal said that this data is also being fed to AI models.
“Today, you can run all the data of customer fitting into AI models and design lens,” he said.
What it means for consent: Under the Digital Personal Data Protection Act (DPDPA), data must be processed strictly for the purpose for which consent was taken. The company conducted 63 lakh eye tests in the April-June quarter. How many of these were AI-enabled eye tests? Did customers seeking a vision check consent to having their biometric data fed into AI models? Is there an opt-out mechanism?
Even if consent is taken for such data collection, there is limited clarity around purpose limitation, specifically, whether data collected for one service can be legally treated as reusable for training AI systems.
Whether Lenskart is using eye test data to design lenses or improve its products and services, a user’s face has already joined a digital ledger, a trace that can’t be easily erased. What if the user’s facial information is stolen or misused? A stolen password can be reset. If credit card credentials get compromised, it can be cancelled. But one cannot reset or revoke the appearance of their cheekbones.
The company’s privacy policy states, “Any photo, facial data or other image that you may share with us, we retain for a period of 5 years from the date of your last use of our website or as long as we continue to provide the services to you or as required by law.”
Though the company claims that facial data is deleted after a fixed timeline, users have limited visibility into how these safeguards are enforced in practice or whether its AI systems continue to feed on parts of this data after deletion.
The threat isn’t theoretical. Biometric data has been stolen in data breaches. In 2023, an investigation by Kolkata’s cybercrime wing revealed that criminals stole people’s biometric details by downloading multiple land deeds from the state’s property registration website. In 2019, an Aadhaar operator’s biometrics were stolen and misused repeatedly.
B by Lenskart poses the same threat as Meta AI glasses: During the earnings call, Lenskart’s CEO said the company has started shipping its AI-powered smart glasses to customers. Dubbed ‘B by Lenskart’, the device can capture audio, images and video and comes with an LED indicator, so the bystander knows the user is recording. Lenskart first announced its AI glasses in Q2 FY26.
“Manufacturing is scaling. A few hundred glasses are shipping daily, and the product is getting better every single week with consumer feedback,” Bansal said.
Lenskart’s smart glasses rollout comes at a time when rival Meta is facing widespread criticism for its AI glasses. Some reports suggested that users were secretly recording women in public while wearing Meta’s smart glasses. The footage was then posted as dating content on social media, where other users subjected the victims to sexualised abuse. India’s data protection law does not require a wearer to ask for consent before recording someone in public. Nor does it mandate that an LED recording indicator remain turned on while filming.
Moreover, MediaNama found that LED-blocking stickers were on sale on Amazon last month. Furthermore, reports surfaced that users were drilling lights out of their Meta AI glasses, allowing them to record people stealthily. There are also widespread concerns about the device’s facial recognition capabilities. Meta is also facing a lawsuit in the US over allegations that its AI glasses recorded sensitive and confidential user data and disclosed it to third-party contractors. These privacy concerns also apply to Lenskart’s smart glasses.
MediaNama has reached out to Lenskart for comments. The story will be updated upon receiving a response.
Questions sent to Lenskart:
1) Lenskart conducted 63 lakh eye tests in the April-June quarter. How many of these were AI-enabled eye tests? During the earnings call, Peyush Bansal said user data is also being fed into AI models to design lenses. Did customers seeking a vision check consent to having their biometric data fed into AI models? Is there an opt-out?
2) Who is responsible when AI misdiagnoses a vision problem?
3) Lenskart has been collecting facial data of users. Unlike personal data, biometric data cannot be reset or revoked once it is connected to a digital ledger. What if this facial information is stolen or misused for identity theft and fraud? What measures has Lenskart implemented to prevent cyber breaches?
4) The company has also seen 80,000 sign-ups for its AI glasses. Does the device still work if the wearer uses an LED-blocking sticker or drills light out of their glasses? Is ‘B by Lenskart’ capable of continuously capturing audio while taking photos of a user’s surroundings? Does it have facial recognition capabilities? What has Lenskart done to address privacy concerns around smart glasses? Multiple reports say, Meta AI glasses were being used to secretly film women in public, then the footage was posted on social media, where they were subjected to abuse.
Key operating metrics:
- Eye Tests Performed: 0.7 crore, up 39.8% YoY
- Eyewear Units Sold: 1.1 crore, a jump of 25.7% YoY
- New Stores Added: 132, up 59% YoY, taking the total active stores to 3,459
Also read:
