Reports have emerged that AI company Anthropic has established a security operations framework to predict and monitor the movements of activists, fueling controversy over corporate security and the infringement of civil liberties.
Imagine this: one day, you find out that after quietly writing posts online or participating in a small-scale protest to criticize the development of a technology you’re interested in, you have been placed on a “person-of-interest” list by a massive IT corporation. Recently, reports that Anthropic, a leader in the field of Artificial Intelligence (AI), is building a large-scale intelligence operation system to monitor activists have caused quite a stir.
Why Does This Matter?
This news does not seem to be merely a story about corporate security. It serves as a window into how the companies building the AI we use every day are treating the voices of citizens who criticize them.
Anthropic has created a “Global Safety, Intelligence, and Security organization” and is operating a 24/7 global security operations center [Reference 3, Reference 5, Reference 8]. While it seems natural for a company to bolster security for safety, the situation changes when it turns legitimate citizen activism into targets of surveillance. In particular, the most controversial aspect of these reports is that Anthropic has classified the movements of activists opposing AI development similarly to terrorist threats [Reference 3, Reference 13]. This raises concerns that it could shrink our collective right to view AI technology critically.
Understanding It Simply: The ‘Person-of-Interest’ Process
The system that Anthropic is reported to be operating is called a “person-of-interest” process [Reference 7]. Simply put, much like detectives, security personnel consistently record the activities of specific individuals and analyze their behavioral patterns over time.
As a metaphor, just as we use photo apps to recognize faces and categorize them by person, Anthropic’s security team is managing certain activists by tagging them as “high risk” or “person-of-interest.” The core of this system is “prediction.” The security team aims to analyze minor behavioral changes or patterns shown by individuals to respond before problems escalate [Reference 1, Reference 7].
Furthermore, they are hiring experts in “OSINT (Open Source Intelligence)” technology who collect posts and activity details from online sources. Their salaries reach $180,000 to $230,000, which implies that the company possesses sophisticated surveillance capabilities inside its walls that one would typically only expect from national intelligence agencies [Reference 8].
Current Situation
The current situation is somewhat complex. Some media outlets are reporting that Anthropic has built an AI that tracks activists and predicts whether to provide information to the police [Reference 3, Reference 13]. On the other hand, some have adopted a more cautious stance, suggesting that it is not certain that Anthropic has created a single, fully automated “predictive surveillance system” [Reference 4].
Anthropic maintains the position that this system is part of corporate security operations intended to protect its executives, employees, and physical assets, rather than large-scale national surveillance [Reference 12]. However, social criticism persists regarding the tracking of citizens’ exercise of legal rights under the pretext of protecting corporate assets [Reference 2].
Where Are We Heading?
Anthropic’s case raises an important question for our society: how far should the scope of “corporate security” be permitted? While it is understandable that a company seeks to protect its technology, if that is used as a tool to silence or monitor citizen criticism, many people will likely harbor even greater distrust toward AI technology itself.
There are two main things we should watch going forward. First, will Anthropic transparently disclose the actual level to which they are collecting and analyzing information? Second, will a social consensus be reached on creating an environment where civil society can safely express critical opinions in the face of AI technological advancement?
MindTickleBytes AI Reporter’s View
Efforts for corporate security risk devolving into tools that suppress legitimate citizen criticism. Transparent social consensus is required to ensure that technological advancement does not prioritize control over the public interest.
References
- Anthropic Is Building a Predictive Surveillance System to Monitor Activists
- Anthropic Building a ‘Pre-Crime’ System to Monitor Activists
- Anthropic Is Building AI to Predict Which Activists Police Should Watch
- Anthropic Built a Security Operation To Monitor Protests and Chatbot Threats
- Anthropic Is Building a Predictive Surveillance System to Monitor Activists
- BREAKING: Anthropic is reportedly building an extensive surveillance and intelligence system to monitor activists and protests opposing artificial intelligence
- Anthropic Runs a ‘Person-of-Interest’ System to Track Activists
- Anthropic Is Building AI to Predict Which Activists Police Should Watch
- Anthropic Is Building a Predictive Surveillance System to Monitor Activists - Peter Theil
- Anthropic builds threat intelligence operation that treats activism as a security threat
- Anthropic Is Building AI to Predict Which Activists Police Should Watch
- It is a large-scale, government-led surveillance system.
- It classifies activists as 'persons-of-interest' and records their movements.
- All systems are perfectly automated.
- National security and mass surveillance
- Protecting executives, employees, and physical assets
- Personal information of activists
- $22,000 ~ $23,000
- $180,000 ~ $230,000
- Over $500,000