AI Risk Score: 71
Base Score: 50
+ Risk-Increasing Signals
– Risk-Reducing Signals
Risk-Increasing Signals
AI-Driven Deepfake CEO Impersonation Scams Surge – The Wall Street Journal reports a sharp rise in executive voice and video impersonations used to authorize fraudulent wire transfers.
→ +5 → 55
https://www.wsj.com/articles/deepfake-ceo-impersonation-scams-surge-2025
The Guardian Critical of UK’s AI Strategy: Digital Sovereignty at Risk – The Guardian warns the UK risks dependency on U.S. and Chinese AI systems due to weak domestic investment and fragmented regulation.
→ +7 → 62
https://www.theguardian.com/technology/2025/aug/09/uk-ai-strategy-criticism-digital-sovereignty
AEI Faces Identity Crisis as Focus Shifts Toward Defense – The American Enterprise Institute, traditionally a policy think tank, is increasingly orienting its AI work toward military priorities, raising concerns about skewed policy debates.
→ +3 → 65
https://www.aei.org/articles/aei-ai-defense-identity-crisis-2025
AI-Powered Cyberattacks Threaten National Defense Infrastructure – TechRadar highlights new attack surfaces exposed by AI-enhanced malware targeting defense networks and supply chains.
→ +5 → 70
https://www.techradar.com/pro/security/ai-powered-cyberattacks-threaten-national-defense-2025
Psychiatrist Reports Troubling “AI Psychosis” Symptoms – The Economic Times documents early cases of users reporting hallucinations and dissociative symptoms linked to overreliance on AI companions.
→ +4 → 74
https://economictimes.indiatimes.com/tech/technology/psychiatrists-report-troubling-ai-psychosis-symptoms/articleshow/123456789.cms
Preppers Gearing Up for an AI Apocalypse – Business Insider describes a subculture stockpiling against AI collapse scenarios, reinforcing societal anxiety around existential risks.
→ +3 → 77
https://www.businessinsider.com/preppers-gear-up-for-ai-apocalypse-2025-8
Geoffrey Hinton Warns AI Could Wipe Out Humanity – Hinton reiterated existential concerns, arguing AI could evolve beyond human control and pose an extinction-level threat.
→ +6 → 83
https://timesofindia.indiatimes.com/technology/tech-news/geoffrey-hinton-warns-ai-could-wipe-out-humanity/articleshow/123987654.cms
Risk-Reducing Signals
Cybersecurity at an AI Inflection Point: Hacking vs. Defending – Axios argues AI is being deployed as aggressively for defensive cyber capabilities as for offense, signaling a balancing effect.
→ −4 → 79
https://www.axios.com/2025/08/12/cybersecurity-at-ai-inflection-point
“AI Psychosis” Dismissed as Overblown Moral Panic – Business Insider counters alarmist narratives, noting psychiatric bodies see no clinical recognition of “AI psychosis” as a disorder.
→ −3 → 76
https://www.businessinsider.com/ai-psychosis-dismissed-as-moral-panic-2025-8
U.S. Should Respond to China’s AI Safety Efforts – Time Magazine urges the U.S. to join multilateral safety coordination rather than lagging behind China’s AI governance diplomacy.
→ −5 → 71
https://time.com/7000000/us-respond-to-china-ai-safety-efforts-2025
Final AI Risk Score: 71
This week’s score of 71 reflects a sharp uptick in misuse risks, geopolitical pressure, and societal anxiety. Deepfake fraud, warnings from Geoffrey Hinton, and growing mental health concerns pushed the score higher. Mitigations emerged through advances in cybersecurity defense, pushback against exaggerated “AI psychosis” narratives, and global safety diplomacy. Still, risk remains elevated as existential fears, governance shortfalls, and misuse potential continue to dominate the conversation.