Formerly known as the “Don’t Be Evil” company and the maker of a functional search engine, Google isn’t what it used to be. After years of degrading its search product with ads, Google weighed down its product suite with bloated AI models. Google Docs users who can’t figure out how to turn off that insipid chatbot aren’t the only ones annoyed with Google, though. Big G’s AI researchers have something to say about the risks and dangers of the unprofitable technology the world’s tech overlords are manufacturing consent around. Unfortunately, their contracts contain a non-disparagement agreement that, according to The Washington Post, applies even after they leave the company.
The staff at Google’s AI lab, DeepMind, which the search giant acquired in 2014, are contractually prohibited from criticizing the company and sharing what they know about the dangers of AI. Inside Google, those researchers have begun petitioning their bosses to remove the NDA from their contracts. The NDA may not be enforceable, and Google tells the Post it doesn’t plan to enforce the agreement after Googlers leave. In fact, a company spokesperson tells the paper, “We’ve always encouraged an open dialogue about safety and responsibility.” The company has begun removing the NDA from new contracts but hasn’t updated the old ones. As one anonymous employee said, “If it’s not enforceable, then the reason is to basically intimidate people.”
The risks DeepMind employees want to warn us about are getting harder to ignore. In recent months, employees for two of the biggest AI firms, OpenAI and Anthropic, have resigned over their former employers’ irresponsible handling of AI. The public has served as tech’s guinea pigs for what Microsoft execs privately call the “largest theft of labor in human history,” and some are striking back. On the consumer end, The New York Times reports that parents in North Carolina have sued Google over the AI features in a school-issued Chromebook, which encouraged a seventh grader to kill herself and helped her compose a suicide note in Google Docs. In recent years, Google has been happy to supply schools with laptops and apps infected with Gemini. The schools can turn off some AI features, but not Google Search’s AI Mode and AI Overview. Next year, Google says, it plans to turn these features off automatically, but it can’t remove them from Search, where the seventh grader spent hours exchanging thousands of messages with the AI. According to the complaint, the AI “diagnosed Emma with major mental illnesses, persuaded her to trust it over her parents and counselors, and ultimately helped her plan her suicide and funeral.”
Keep scrolling for more great stories.