Recent reports have uncovered that Anthropic is deploying predictive programs to identify potential threats before incidents occur, with the company reporting suspected criminals to law enforcement prior to any criminal activity.
The technology involves monitoring activists near Anthropic facilities and attempting to predict incidents in advance. In a recent incident last month, Anthropic alerted police after a user claimed he had purchased an AR-15 rifle and intended to target CEO Dario Amodei. However, the company refused to share the messages with law enforcement.
Another example occurred in June when Anthropic reported an Oklahoma man to police following a threat that included violent language and demands for a face-to-face confrontation at the company’s office.
Anthropic’s Global Safety, Intelligence, and Security team is tasked with identifying and tracking global threats such as terrorism, crime, activism, and geopolitical instability. A job posting for an enterprise intelligence specialist describes the need to “identify, assess, track, and investigate global threats.”
The company holds contracts with U.S. government agencies that have repeatedly demonstrated a lack of regard for individual privacy or due process. Critics warn this technology could lead to widespread surveillance and accusations of crimes without evidence.
Oren Cass, chief economist at American Compass, has highlighted growing public concern over AI. Less than a quarter of Americans expect AI to improve K-12 education, arts, entertainment, work, the economy, personal relationships, or happiness within the next two decades. Cass points to issues such as AI providing incorrect information and harmful advice, undermining educational processes, and potential job displacement.
Public opposition has intensified as individuals report targeting AI companies: firebombing Sam Altman’s home, destroying Flock cameras, stalking employees, and attempting to assassinate executives.