Meta is being looked at closely after ending its contract with a Kenyan AI training firm because of claims that workers were shown graphic content on smart glasses and worries about openness.
People are asking Meta to explain
People are asking Meta to explain why it broke a costly deal with a company that it was using to train AI. Reports say the decision was made soon after some Kenyan employees of the company said they had to look at graphic content caught by Meta smart glasses.
Swedish media said that employees of the company Sama had seen people wearing the smart glasses doing sexual acts and even going to the bathroom with them on. Less than two months later, Meta stopped its deal with Sama, which is said to have put 1,108 people out of work.
Later, Meta made it clear that the choice was made because Sama couldn’t meet the standards it set. Sama denied this claim. A group of workers says that the choice was made because employees spoke out in public.
Swedish newspapers Svenska Dagbladet and Göteborgs-Posten released stories from workers who did not want to be named but were asked to look over a lot of videos that were shot with Meta smart glasses.
One worker told a Swedish newspaper, “We see everything from living rooms to naked bodies.”
At the time of publishing, Meta admitted that hired workers may have had to look over information recorded on its smart glasses when users sent it to Meta AI. It also said that this was a regular way to make things better for customers.
The UK Information Commissioner’s Office was interested in the findings and wrote to Meta to say that the story was “concerning.” At the same time, Kenya’s data security commissioner began a probe into the issue.
Meta told the BBC about the layoffs, “Last month, we stopped working with Sama while we looked into these claims.” They are important to us. Users can only see their own photos and movies. People look over AI material to make the product work better, and users give their clear permission for this to happen.
Concerns have also been raised about the wrong use of smart glasses and the fact that content reviewers can see private user videos, as a BBC study explained.
This is not the first time something like this has happened. Former workers who used to moderate content for sites like Facebook also complained and took legal action, and some said they were exposed to similarly violent material.
read also
- China blocks Meta’s $2 billion AI deal, shocking the tech industry worldwide
- Meta–YouTube ruling explained What it means for social media users and platforms
Naftali Wambalo of the Africa Tech Workers Movement told the BBC
Naftali Wambalo of the Africa Tech Workers Movement told the BBC that he had talked to workers who were involved in Meta’s smart glasses deal. Wambalo is a petitioner in a case that is still going on. He said the project stopped because the company didn’t want workers to say that people sometimes look over the content that the devices record, saying it was a matter of “secrecy standards.”
Meta has said in the past that its terms of service tell users that human review may happen.
Mercy Mutemi, the executive head of the Oversight Lab and a lawyer for the petitions, said that Meta’s answer should be a message to the Kenyan government.
I am a versatile content writer from the MP region, covering politics, business, crime, current affairs, entertainment, video games, and sports with clear insights, engaging analysis, and timely, reader-focused updates.
