Public investigations

Published investigations explain what the reviewed public record supports, what remains uncertain, and where the evidence comes from.

• Did ChatGPT Help the Tumbler Ridge Shooter? — OpenAI acknowledges that an automated system flagged an account linked to the attacker in June 2025, that people reviewed and banned it, and that police were not notified under the company’s then-current threshold. OpenAI later found a second account. The RCMP now says companies receiving requests complied, while public reporting says OpenAI provided chat logs to law enforcement; the public record does not establish their full scope or make them public. Plaintiffs allege chatbot assistance, but DarkSignals has not authenticated the underlying conversations. The evidence establishes a consequential detection-and-escalation failure; it does not establish material assistance, legal causation or preventability.
• Will AI Destroy Humanity? — No publicly documented AI system can independently cause human extinction today. Current systems remain dependent on human-controlled infrastructure, but they already amplify cyber, biological, fraud and manipulation risks. Substantially more capable systems could create a genuine loss-of-control risk, although the evidence does not establish that outcome or support a reliable probability. DarkSignals therefore assesses AI extinction as a low-probability, high-consequence risk, with moderate confidence.
• Will Russia Attack Lithuania — An open-source assessment of Russian intent, capability, warning indicators and deterrence affecting Lithuania, Belarus, Kaliningrad and NATO's eastern flank.