Long-serving employees at AI safety company Anthropic are reportedly purchasing remote land in the US as a potential refuge if artificial intelligence goes catastrophically wrong, according to the Wall Street Journal.
The purchases reflect deep connections between Anthropic's leadership and the Effective Altruism movement, a philosophical framework focused on reducing existential risks. The Journal's report traces these concerns back over a decade to a network of researchers in the Bay Area who have been modeling worst-case AI scenarios.
Anthropric, founded in 2021 by former OpenAI executives including Dario and Daniela Amodei, has positioned itself as an AI safety-focused company. Its public mission centers on developing AI systems that are beneficial, harmless, and honest. The company has raised billions in funding from major investors including Google and Salesforce.
The land purchases suggest some of the company's veterans are taking their concerns about AI risks beyond corporate strategy and into personal preparation. The remote locations would theoretically provide isolation from potential harms stemming from advanced AI systems.
Effective Altruism, the movement underlying these concerns, emphasizes addressing existential and global catastrophic risks as a moral priority. Within tech circles, EA-aligned researchers have long debated scenarios where artificial superintelligence could pose dangers to humanity.
While doomsday preparation might seem extreme, the practice reflects genuine anxiety within parts of the AI safety community. These concerns aren't universally shared—many AI researchers and industry leaders view catastrophic AI risks as speculative or distant.
Anthropric's public stance emphasizes building safer AI systems rather than preparing for worst-case outcomes. The company's research focuses on AI alignment and interpretability. However, the personal actions of its employees reveal underlying philosophical positions that may extend beyond the company's official messaging.
The report doesn't specify how many employees are involved in the land purchases or identify individuals by name. It remains unclear whether this represents a broader trend at Anthropic or is limited to a smaller subset of staff members with particularly strong doomsday concerns.
Bill Gates said government AI safety regulations won't disadvantage the US against China, disagreeing with President Trump's light-touch regulatory stance on artificial intelligence.
Meta has blocked Brazilian President Luiz Inácio Lula da Silva's Facebook page and campaign advertisements just two weeks before the election. The action has sparked immediate controversy over content moderation decisions during a critical political period.
OpenAI leadership was aware that mass book piracy violated copyright law, according to court filings in a lawsuit by the Authors Guild. Internal communications reveal the company feared negative attention on platforms like Hacker News.
Meta will accept advertising for Alex Gibney's upcoming documentary about Elon Musk, reversing an earlier position. The announcement marks a shift in the social media platform's approach to the film's promotion.