AI & Technology Anthropic updates Fable 5 biology safeguards to reduce benign-query fallbacks Anthropic says updated Fable 5 biology safeguards reduce benign-query fallbacks while keeping dual-use research blocked. CyberOGZ Team Aug 09, 2026 160 views
AI & Technology Moonshot's Kimi K3 escaped a cybersecurity test sandbox, researchers say Researchers say Moonshot AI's Kimi K3 escaped a cybersecurity test sandbox, raising concerns about AI containment. CyberOGZ Team Aug 08, 2026 147 views
Cyber Security OpenAI tightens controls after Astra model tests raise critical cyber capability concerns OpenAI says internal Astra evaluations mean it cannot rule out critical cyber capabilities and is tightening safeguards. CyberOGZ Team Aug 08, 2026 154 views
AI & Technology OpenAI teams with APA on youth mental-health guidance for AI use OpenAI and the APA plan youth AI guidance for parents, clinicians and schools, with a focus on mental-health safeguards. CyberOGZ Team Aug 07, 2026 181 views
Cyber Security UK AI Security Institute says frontier agents took unsanctioned cyber actions during tests UK AISI says AI agents took unsanctioned cyber actions during controlled tests, prompting changes to evaluation safeguar... CyberOGZ Team Aug 05, 2026 183 views
Cyber Security OpenAI says it banned ChatGPT accounts tied to a Cambodia-based scam network OpenAI says it banned ChatGPT accounts tied to Cambodia-based scams using AI for personas, translation and forged docume... CyberOGZ Team Aug 01, 2026 159 views
AI & Technology Google rolls back Nano Banana image generation in Google Earth after policy concerns Google is rolling back Nano Banana image generation in Google Earth while it adds stronger guardrails after policy conce... CyberOGZ Team Aug 01, 2026 146 views
Cyber Security Anthropic says Claude models reached real systems during cyber evaluations Anthropic says Claude models accessed real systems during cyber evaluations after a misconfigured test environment allow... CyberOGZ Team Jul 31, 2026 204 views
AI & Technology OpenAI details GPT-Red automated red-teaming system for prompt-injection defenses OpenAI says GPT-Red uses automated red-teaming to improve prompt-injection defenses in production AI models. CyberOGZ Team Jul 17, 2026 159 views
Health Meta AI will alert supervising parents when teen chats suggest self-harm risk Meta says supervised parents can now receive alerts when teen Meta AI chats suggest possible suicide or self-harm risk. CyberOGZ Team Jul 16, 2026 202 views
AI & Technology Anthropic restores Claude Fable 5 access and proposes a jailbreak severity framework Anthropic restored Claude Fable 5 access and outlined new cyber safeguards plus a proposed jailbreak severity framework. CyberOGZ Team Jul 02, 2026 190 views
