{"slug": "anthropic-built-a-security-operation-to-monitor-protests-and-threats-around-its", "title": "Anthropic Built a Security Operation To Monitor Protests and Threats Around Its Executives", "summary": "Anthropic has built an extensive internal security operation that monitors protests near its executives and offices, and has reported users to police over threats made via its Claude chatbot, according to a report by The American Prospect published on 9 September. The company's Global Security Operations Center Manager, Keon Ellison, described using risk-detection firm Samdesk to reroute an executive away from a protest with 60 minutes' notice, and in one case, Anthropic reported a man to San Francisco police after he told Claude he had an AR-15 and CEO Dario Amodei 'in his sights', but withheld the actual messages from police. A job listing for an enterprise intelligence specialist with a salary of $180,000 to $230,000 indicates the company is expanding its security work to track threats including 'geopolitical instability, terrorism, crime, activism, nation-state targeting of the AI sector, and emerging security trends'.", "body_md": "# Anthropic Built a Security Operation To Monitor Protests and Threats Around Its Executives\n\n## Inside Anthropic's extensive security operations and its approach to handling chatbot threats\n\n[Anthropic](https://www.ibtimes.co.uk/anthropic-security-concerns-allied-universal-dispute-1816623) has built out an extensive internal security operation that monitors protests near its executives and offices, according to a report by The American Prospect. The firm has also reported users to police over messages sent to its Claude chatbot, in at least one case without showing police the actual messages.\n\nThe revelations come from a podcast interview with two of Anthropic's senior security officials, alongside a recent job listing and past comments the company gave to other outlets. Anthropic did not respond to The American Prospect's request for comment on the story, published on 9 September.\n\n## How Anthropic Tracks Protests Near Its Executives\n\nIn the interview, Anthropic's Global Security Operations Center Manager, Keon Ellison, described a case where the company used intelligence from risk-detection firm Samdesk to reroute an executive travelling into a city where a protest was planned.\n\n'Samdesk gave us about 60 minutes of advance notice that the protest organisers had moved the timeline,' Ellison said. 'That extra hour was critical. Without it, our executives would have departed their meetings, they would have ran right into the heart of the disruption.'\n\nEllison said the company used the tip to send the executive to a hotel's service entrance instead. He described the outcome as a situation 'mitigated through early detection through Samdesk and giving us that information'.\n\n## Reported To Police, but Messages Withheld\n\nAnthropic has also begun making reports to police departments over threats linked to its Claude chatbot.\n\nThe company told The Wall Street Journal in July that it tracks 'concerning behaviour over time through a person-of-interest process, allowing us to catch escalation patterns early', and the paper reported that several people later involved in police incidents had already been flagged by Anthropic's security team.\n\nIn one case reported by The San Francisco Standard last month, Anthropic reported a man to San Francisco police after he told Claude he had bought an AR-15 rifle and had chief executive Dario Amodei 'in his sights'. The man later told the Standard he was 'just f**king around'.\n\nAn Anthropic employee did not show the man's actual messages to police, citing company policy as the reason for withholding them. The incident raises questions about how police assess reported threats when they are not shown the original communication.\n\n## Job Listing Points To Wider Intelligence Push\n\nA job listing for Anthropic's Global Safety, Intelligence, and Security team indicates that the company's security work extends beyond executive protection.\n\nThe listing, for an enterprise intelligence specialist with a salary range of $180,000 to $230,000, calls for investigating and tracking threats including 'geopolitical instability, terrorism, crime, activism, nation-state targeting of the AI sector, and emerging security trends'.\n\nZach Melvin, Anthropic's Security Operations Manager, said in the same podcast that the aim was 'transforming operations from reactive information gathering to proactive and predictive and preventative threat engagement and management'.\n\nThe expansion comes as Anthropic pursues renewed defence business, having previously clashed with the [Pentagon](https://www.ibtimes.co.uk/ai-giants-pentagon-military-contracts-1818594) over military use of its tools.\n\nThe company is now seeking to restart military contracts, according to The American Prospect, while its national security hiring has expanded.\n\nThe reporting does not establish that Anthropic has built a single automated 'predictive [surveillance](https://www.ibtimes.co.uk/ai-expert-warns-totalitarian-risks-surveillance-technologies-1814730)' system.\n\nInstead, it points to a broader security and intelligence operation involving protest monitoring, threat assessments and proactive reporting. Anthropic's privacy policy allows the company to report threats of violence in chats to law enforcement.\n\nThe developments raise questions about how companies handle potentially threatening chatbot conversations and what information they provide to police after making such reports. They also come as Anthropic increases its government and national security engagement, including lobbying on issues such as defence procurement and cybersecurity.\n\n© Copyright IBTimes 2026. All rights reserved.", "url": "https://wpnews.pro/news/anthropic-built-a-security-operation-to-monitor-protests-and-threats-around-its", "canonical_source": "https://www.ibtimes.co.uk/anthropic-security-measures-monitoring-protests-chatbot-threats-1818850", "published_at": "2026-09-09 22:03:15+00:00", "updated_at": "2026-09-09 22:16:07.723751+00:00", "lang": "en", "topics": ["ai-policy", "ai-ethics", "ai-products"], "entities": ["Anthropic", "Keon Ellison", "Samdesk", "Dario Amodei", "The American Prospect", "The Wall Street Journal", "The San Francisco Standard", "Zach Melvin"], "alternates": {"html": "https://wpnews.pro/news/anthropic-built-a-security-operation-to-monitor-protests-and-threats-around-its", "markdown": "https://wpnews.pro/news/anthropic-built-a-security-operation-to-monitor-protests-and-threats-around-its.md", "text": "https://wpnews.pro/news/anthropic-built-a-security-operation-to-monitor-protests-and-threats-around-its.txt", "jsonld": "https://wpnews.pro/news/anthropic-built-a-security-operation-to-monitor-protests-and-threats-around-its.jsonld"}}