You remember the pitch. We were exhausted by the Silicon Valley hustle bros racing to build unchecked, world-ending tech. So, when Anthropic stepped onto the stage, we breathed a sigh of relief. They were the “safe” AI company. The adults in the room. They cared about alignment, ethics, and not destroying humanity.
But we’ve been played for fools.
It turns out the “responsible” AI company is building a predictive surveillance system to monitor activists. Yes, the exact same dystopian tech we usually associate with overtly authoritarian defense contractors like Palantir. But because it’s wrapped in a bow of “safety” and “alignment,” we’re supposed to look the other way. Safety is just a brand. Power is the product.
Let’s drop the PR speak. Anthropic isn’t building a shield to protect you; they’re building a cage to contain you. When an AI firm builds tools to predict and monitor political dissent, they aren’t “protecting democracy.” They are building the exact infrastructure a future authoritarian will use to crush it. The most dangerous authoritarian isn’t a dictator in a military uniform; it’s a CEO in a Patagonia vest telling you it’s for your own good.
You might think, “Well, I’m not an activist, so this doesn’t affect me.” That’s the trap. The tools being built today will outlast the companies and leaders who created them. Today, it’s the activists. Tomorrow, when the next government decides your dissent is inconvenient, it’s you. The tech doesn’t expire when the administration changes. The infrastructure remains.
We wanted to believe there were good guys in the AI arms race. But there are no good guys—only vendors. Palantir doesn’t pretend to be your friend, which makes them easier to distrust. Anthropic smiles, talks about ethics, and sells the exact same surveillance machinery under the guise of keeping us secure. Tools built to protect the system will always eventually be used to protect the system from you. Stop trusting the label. Start fearing the infrastructure.
FAQ
Q: Doesn't Anthropic need to monitor for threats to keep AI safe?
A: No, this isn't about monitoring their models for flaws; it's about monitoring *people* and their political activities. Predictive surveillance of activists has nothing to do with AI alignment and everything to do with social control.
Q: How does this actually affect a normal person?
A: It shifts the baseline of acceptable monitoring. Once 'ethical' companies normalize predictive surveillance, it becomes standard infrastructure. When the political winds shift, that infrastructure is already in place to target anyone.
Q: Is this really worse than what Palantir does?
A: Yes. Palantir faces public scrutiny and backlash. Anthropic uses its 'ethical' halo to launder the exact same technology, making it palatable to the public and easier for governments to adopt without pushback.