# Anthropic is building a protest-prediction machine

> An AI lab known for safety talk is now watching its critics and trying to guess their next move.
- Title: Anthropic is building a protest-prediction machine · Futurism
- Summary: An AI lab known for safety talk is now watching its critics and trying to guess their next move. The company warning loudest about AI risk is quietly…
- Keywords: anthropic, surveillance, ai protest, privacy, security, science, technology, building, protest-prediction, machine, Futurism
- Source: Futurism — https://futurism.com/artificial-intelligence/anthropic-surveillance-predictive-system-spy-ai-activists
- Author: Joe Wilkins
- Published: 2026-09-10T14:43:14+00:00
- Read time: 1 min
- Topics: anthropic, surveillance, ai protest, privacy, security, science, technology
## A lab watching its critics
As opposition to AI grows, Anthropic is standing up a surveillance operation aimed at tracking dissenters and forecasting when the next wave of protest hits.
## How the story surfaced
The American Prospect broke it after talking to Anthropic senior security officials and combing through the company's job postings to map the scope of the operation.
## Pre-crime for protests
The watch list goes beyond demonstrations near Anthropic offices. Security staff describe a pre-crime posture aimed at spotting potential civil unrest before it forms.
## The 60 minute save
Security manager Keon Ellison described a vendor, Samdesk, flagging that organizers had moved a protest up an hour, letting executives avoid walking into it.

> Without it our executives would have departed their meetings, they would have ran right into the heart of the disruption.
## Reported to police, no logs
A San Francisco man told Claude he had a rifle and CEO Dario Amodei in his sights. Anthropic called the cops, then declined to hand over the chat logs as evidence.
## Reactive to predictive
The stated goal is to stop responding to events and start anticipating them, the same posture companies use to protect what they call high-value targets.

> Transforming operations from reactive information to gathering proactive and predictive and preventative threat engagement and management.
## Key takeaway

The company warning loudest about AI risk is quietly building the kind of predictive surveillance it says it fears.