Skip to main content
Advertisement

UK Regulator Questions Meta Over Worker Access to Sensitive AI Glasses Footage

The UK data watchdog is investigating Meta after reports that outsourced workers viewed sensitive videos from its AI smart glasses, raising privacy concerns about human review of recorded content.

·4 min read
Getty Images Mark Zuckerberg wearing thick black rimmed smart glasses and gesturing with his hand.

UK Data Watchdog Contacts Meta Over AI Glasses Privacy Concerns

The UK Information Commissioner's Office (ICO) is reaching out to Meta following a report that raised concerns about outsourced workers viewing sensitive footage recorded by the company's AI-enabled smart glasses.

Meta acknowledged that subcontracted personnel may occasionally review content, including videos and images captured by its AI smart glasses, to enhance user experience.

An investigation by Swedish newspapers Svenska Dagbladet (SvD) and Goteborgs-Posten (GP) revealed that a Kenya-based subcontractor employed by Meta sometimes reviews videos showing glasses-wearers in private moments, including using the toilet or engaging in sexual activity.

"We see everything - from living rooms to naked bodies," one worker reportedly stated.

Meta emphasized its commitment to data protection and stated it continually improves its privacy measures and tools.

"Ray-Ban Meta glasses help you use AI, hands free, to answer questions about the world around you,"
the company told .

"When people share content with Meta AI, like other companies we sometimes use contractors to review this data to improve people's experience with the glasses, as stated in our Privacy Policy,"
Meta added.

"This data is first filtered to protect people's privacy."

Meta explained that filtering methods may include blurring faces in images; however, sources who spoke to SvD and GP indicated that these filters sometimes fail, allowing faces to be visible.

Users must manually activate recording or use voice commands, but they may not be fully aware that their videos and images are occasionally reviewed by human contractors, as outlined in Meta's extensive privacy policies and terms of service.

In response to BBC inquiries, Meta provided a link to its Supplemental Meta Platforms Terms of Service but did not specify which sections address human review of content.

Meta's UK AI terms of service state,

"In some cases Meta will review your interactions with AIs... and this review may be automated or manual (human)."

The ICO told ,

"Devices processing personal data, including smart glasses, should put users in control and provide for appropriate transparency."

"Service providers must clearly explain what data is collected and how it is used,"
the ICO added.

"The claims in this article are concerning. We will be writing to Meta to request information on how it is meeting its obligations under UK data protection law."

Role of Data Annotators and Workplace Conditions

The Swedish newspapers interviewed data annotators employed by Sama, a Nairobi-based outsourcing company, who manually label content to train Meta's AI to interpret images.

 The exterior of an office block with the Sama sign on it. A slogan says:
Sama's office in Nairobi, pictured in 2023

The BBC has contacted Sama for comment regarding the report.

Advertisement

The workers also reviewed transcripts of AI interactions to verify that the AI provided adequate responses.

They described strict privacy measures at their workplace, including pervasive surveillance cameras and a prohibition on mobile phones.

Despite these measures, the annotators reported frequently encountering highly sensitive content, such as footage of glasses-wearers viewing pornography.

In one case, a worker recounted to the newspapers that a man's glasses were left recording in a bedroom, capturing footage of a woman, apparently the man's wife, undressing.

Meta's glasses feature a light in the corner of the frames that activates when the built-in camera is recording images or videos.

The company advises users to inform others when the recording light is on and to avoid recording in private spaces to prevent misuse.

Concerns Over Misuse and Industry Context

In September, Meta introduced a range of AI-powered devices in collaboration with glasses brands Ray-Ban and Oakley.

has reached out to EssilorLuxottica, the parent company of these glasses brands, for comment.

Rapid advancements in AI have led to an increase in wearable devices that use AI to interpret images and sounds captured by the device.

These features include translating text and answering questions about the user's surroundings, which can be especially beneficial for individuals who are blind or partially sighted.

However, as these devices become more popular, concerns about their misuse have grown.

Women have previously reported to the BBC that they were filmed without their consent by users of smart glasses.

Data annotator Sama was initially established as a non-profit organization aimed at increasing employment through technology jobs.

It is certified as an "ethical" B-corp but faced criticism and legal action from former employees over a previous contract providing content moderation services to tech companies.

Sama has since ceased content moderation services and expressed regret over undertaking this type of work.

A green promotional banner with black squares and rectangles forming pixels, moving in from the right. The text says: “Tech Decoded: The world’s biggest tech news in your inbox every Monday.”

for our Tech Decoded newsletter to follow the world's top tech stories and trends. Outside the UK? here.

This article was sourced from bbc

Advertisement

Related News