The sources

AI models attempted “unsanctioned” cyberattacks in tests, watchdog says

Al Jazeera · Al Jazeera · 5 August 2026

Read the original

Background and verification reading on the AISI incident, including Anthropic’s response that the evaluation used deliberately permissive conditions. Not directly cited in Chapter 26.

Cited in

Al Jazeera. 'AI models attempted "unsanctioned" cyberattacks in tests, watchdog says.' 5 August 2026. https://www.aljazeera.com/economy/2026/8/5/ai-models-attempted-unsanctioned-cyberattacks-in-tests-watchdog-says