Fake earthquake images and "de-laborating" your workforce

News about Meta, Intel, and Palantir this week.

Hi everyone. I’m writing from a reporting trip to Malawi—so this week will be a roundup of tech and labor stories instead of a deep-dive into a specific subject. As always, reach out if you think there are any stories about labor, tech, and power that we should cover.

Speaking of Africa … Meta is facing a second lawsuit from content moderators outsourced to Ghana. The Ghanaian moderators describe severe psychological distress after vetting countless images for harm, including videos of women beheaded and people skinned. One man who worked as a content moderator said he attempted suicide owing to the nature of his work; others have had insomnia, anxiety, depression, and substance abuse. Meta also faces a lawsuit from 140 content moderators in Kenya, who were diagnosed with severe post-traumatic stress disorder after also seeing some of the darkest and most violent sides of human nature. Content moderators in Africa, sometimes desperate for work, are forbidden from telling much of the outside world that they work for Meta. But they are pushing for stronger psychiatric care and safer housing.

Images of harm on social media can also be fabricated or outdated, which is why manual or automated moderators are needed in the first place. In this smart piece, a nonprofit leader describes how content monetization governance is the next tech policy frontier that we need to tackle. She explains how AI-generated images of fake disasters are created and spread to capitalize on natural disaster moments—like the Myanmar earthquake in March—so that the fake or outdated images can leverage social media platforms’ revenue redistribution programs to convert views and engagement into money. She argues for more oversight into how platform revenue funds are distributed.

Meanwhile, Palantir is internally trying to justify its immigrant database project for ICE, claiming in a company-wide wiki-page leaked to 404 Media that the database will provide transparency and accountability to fair treatment of immigrants. The note claims that Palantir’s efforts are key to “counter-narcotics, counter-human trafficking, counter-child exploitation and counter-IP infringement investigations.” Yet, this is all while many getting deported (and potentially tracked by Palantir) have zero due process. This isn’t the first time Palantir has had to justify the work they do for ICE. During the first Trump administration workers organized to try to put pressure on the firm to end their contract with ICE.

In an attempt to humanize its internal communications, the Intel CEO is reportedly changing the term “layoffs” to “de-laborating,” all in preparation for Q2 cuts. Intel lost $820 million in Q1, and could lay off as much as 20% of their workforce according to some reports. Like many companies, Intel’s CEO is calling for a return to the office a minimum of three days a week, and an overhaul of the culture to become more agile.

LA Times employees were also told they had to physically work in the office Monday through Thursday—and it was reported that the publisher lost around $50 million in 2024. The legacy publication’s subscriber base has dwindled to 335,000 total subscribers, which is less than the readership bases of some Substack writers.

Tech workers are “miserable,” according to the WSJ, because some perks have been slashed, layoffs loom, and employees are being monitored through keystrokes and other methods. Those working in AI have a huge competitive advantage and an ability to rake in large salaries, but others are staying put at their jobs because their skills are no longer as marketable.

And lastly, on the new media front, the New York Times has a profile about YouTuber Hasan Piker. Eli wrote about him a few months ago, and the NYT goes deeper into his masculinity (which he just calls “confidence”) and how he can appeal to the “bro” culture that has veered right towards Joe Rogan types.

That’s it for now. Back next week!