Meta terminates contract after AI trainers report seeing intimate footage from Ray-Ban smart glasses

Meta Ends Contract Following Worker Allegations

Meta terminated its contract with Sama, a Kenya-headquartered outsourcing firm, less than two months after workers reported viewing sensitive and explicit footage captured by Ray-Ban smart glasses users 1

1

. The decision affected 1,108 workers who performed data annotation work for Meta's AI systems 2

2

. In February, Sama employees told Swedish newspapers Svenska Dagbladet and Göteborgs-Posten, along with Kenya-based freelance journalist Naipanoi Lepapa, that they witnessed footage of people performing intimate acts, changing clothes, having sex, and using the toilet 1

1

.

A Meta spokesperson stated the company "decided to end our work with Sama because they don't meet our standards," though the company has not specified which standards were breached 3

3

. Sama rejected this characterization, asserting it "consistently met the operational, security, and quality standards required across our client engagements, including with Meta" and was never notified of any failure to meet those standards 3

3

.

Allegations of Retaliation Against Whistleblowers

Naftali Wambalo of the Africa Tech Workers Movement alleges Meta's decision represents retaliation against workers who spoke out about user privacy concerns. "What I think are the standards they are talking about here are standards of secrecy," Wambalo told the BBC 5

5

. Workers received only six days' notice before their positions were made redundant on April 16 4

4

. Meta has not addressed the retaliation allegation directly, but emphasized it takes the claims seriously and that "photos and videos are private to users" 3

3

.

The workers were tasked with labelling images and video to train Meta's AI models, a process requiring them to view, categorize, and describe whatever the cameras had recorded 4

4

. One anonymous employee was quoted saying they "are just expected to carry out the work" even when viewing private footage, with limited ability to question the content 1

1

.

How AI Data Collection Works Behind the Scenes

Meta confirmed it sometimes shares content that glasses owners provide to the Meta AI generative AI chatbot with contractors to improve product performance, stating such data is first filtered to protect privacy by blurring faces in pictures 1

1

. However, the Swedish investigation revealed the footage was not anonymized before review, and workers could see faces, bodies, and personal documents 4

4

.

Ray-Ban smart glasses display a light when recording, but Sama workers reported it appeared some users remained unaware their glasses were recording. "People can record themselves in the wrong way and not even know what they are recording," one employee stated 1

1

. The lights can reportedly be disabled, and users have discovered tricks to cover them 5

5

. Meta sold more than seven million pairs of Ray-Ban smart glasses in 2025, more than tripling the previous year's volume 4

4

.

Regulatory Investigations and Legal Action

The revelations have triggered regulatory investigations and a class-action lawsuit against Meta. In March, a complaint was filed in the US District Court for the Northern District of California against Meta and Luxottica of America, a subsidiary of EssilorLuxottica, Ray-Ban's parent company 1

1

. The complaint accuses Meta of breaking state consumer protection laws and seeks damages, punitive penalties, and an injunction requiring changes to prevent consumer deception 1

1

.

The UK's Information Commissioner's Office sent Meta a letter about the "concerning" reports, stating that "devices processing personal data, including smart glasses, should put users in control and provide appropriate transparency" 1

1

. The Office of Kenya's Data Protection Commissioner announced it was investigating "privacy concerns raised in relation to the Ray-Ban Meta glasses and the processing of personally identifiable information for the training of Meta AI" 1

1

.

Ethical Implications for AI Industry Labor Practices

The controversy highlights broader ethical implications regarding AI industry labor practices and content moderation. Sama, founded in 2008 as a social enterprise with the mission of providing dignified digital work to low-income communities, has operations in Kenya, Uganda, and India 4

4

. This isn't the first time a contract with Meta has soured for Sama. An earlier deal to moderate Facebook posts attracted criticism and legal action by former employees who described exposure to graphic, traumatizing content 3

3

. Sama is also the firm that OpenAI contracted to train ChatGPT, where Kenya-based workers were paid less than $2 a day to filter distressing content 2

2

.

Mercy Mutemi, a lawyer representing petitioners and executive director of the Oversight Lab, warned that Meta's statement should alert the Kenyan government. "We've been told that this is our entry route into the AI ecosystem. This is a very flimsy foundation to build your entire industry on," she told the BBC 3

3

. The case raises questions about user consent and whether Meta's privacy policy adequately discloses that human reviewers may view intimate footage from smart glasses. While Meta's policy states users who opt into AI training allow footage processing, it does not emphasize the human layer between camera and algorithm 4

4

.

What This Means for AI Wearables and Privacy

The allegations add to perceptions that Meta's AI wearables function as "pervert glasses," allowing users to discreetly record people without consent or knowledge 5

5

. As Meta expands its product line to include prescription models designed to reach billions who buy corrective eyewear, converting what was a novelty into something closer to a default, scrutiny over data privacy breach concerns will likely intensify 4

4

.

Wearers have been documented using the devices inside courtrooms, during police operations, and in classrooms during exams 2

2

. Apple is reportedly testing up to four smart glass designs to compete with Meta 2

2

, suggesting the market for AI wearables will expand even as AI ethics debates continue.