I’m taking a data security posture management (DSPM) certificate course, and I’m having a real hard time with the the notion that the top maturity level is basically “add #AI for continuous management!”

The data destruction piece literally says, “At Level 5, organizations fully eliminate manual intervention in data destruction, relying on AI-powered automation to refine policies, enforce compliance, and manage security risks dynamically.”

Yeah, that sounds nice, but then I look at the stories like how #claude AI deleted the #pocketOS database, or how #OpenClaw deleted the mailbox of of the Director of Safety and Alignment at #Meta.

We need humans in the loop. And we need to ask ourselves what benefits we’re getting by introducing non-deterministic LLMs and AIs into #compliance processes that should be precise.

CVE Alert: CVE-2026-45223 - openclaw - crabbox - https://www.redpacketsecurity.com/cve-alert-cve-2026-45223-openclaw-crabbox/

#OSINT #ThreatIntel #CyberSecurity #cve-2026-45223 #openclaw #crabbox

CVE Alert: CVE-2026-8305 - n/a - OpenClaw - https://www.redpacketsecurity.com/cve-alert-cve-2026-8305-n-a-openclaw/

#OSINT #ThreatIntel #CyberSecurity #cve-2026-8305 #n-a #openclaw

CVE Alert: CVE-2026-45006 - OpenClaw - OpenClaw - https://www.redpacketsecurity.com/cve-alert-cve-2026-45006-openclaw-openclaw/

#OSINT #ThreatIntel #CyberSecurity #cve-2026-45006 #openclaw #

CVE Alert: CVE-2026-45224 - openclaw - crabbox - https://www.redpacketsecurity.com/cve-alert-cve-2026-45224-openclaw-crabbox/

#OSINT #ThreatIntel #CyberSecurity #cve-2026-45224 #openclaw #crabbox

CVE Alert: CVE-2026-45004 - OpenClaw - OpenClaw - https://www.redpacketsecurity.com/cve-alert-cve-2026-45004-openclaw-openclaw/

#OSINT #ThreatIntel #CyberSecurity #cve-2026-45004 #openclaw #

CVE Alert: CVE-2026-45001 - OpenClaw - OpenClaw - https://www.redpacketsecurity.com/cve-alert-cve-2026-45001-openclaw-openclaw/

#OSINT #ThreatIntel #CyberSecurity #cve-2026-45001 #openclaw #

CVE Alert: CVE-2026-44995 - OpenClaw - OpenClaw - https://www.redpacketsecurity.com/cve-alert-cve-2026-44995-openclaw-openclaw/

#OSINT #ThreatIntel #CyberSecurity #cve-2026-44995 #openclaw #

The Guardian | AI-powered hacking has exploded into industrial-scale threat, Google says by Aisha Down and Dan Milmo

AI generated summary, Read the full article for complete information.

In a recent Google threat‑intelligence report, AI‑powered hacking is already an industrial‑scale threat, with criminal groups and state‑linked actors from China, North Korea and Russia leveraging commercial large‑language models such as Gemini, Claude and OpenAI tools to accelerate, scale and sophisticate attacks—including faster malware development, persistence, and zero‑day exploitation. Google’s chief analyst John Hultquist says the “AI vulnerability race” has begun, noting that AI lets threat actors test operations, refine exploits and launch mass‑exploitation campaigns, as illustrated by a group on the brink of using a non‑Mythos LLM for a large‑scale zero‑day attack. Anthropic’s decision to withhold its Mythos model after it uncovered pervasive zero‑day flaws underscores the danger, while experts like UCL’s Steven Murdoch warn AI is reshaping vulnerability discovery. Meanwhile, the Ada Lovelace Institute cautions that public‑sector productivity gains touted for AI may rest on untested assumptions, urging more rigorous, long‑term evaluation of AI’s real impact.

Read more: https://www.theguardian.com/technology/2026/may/11/ai-powered-hacking-industrial-scale-threat-three-months-google

#Google #Anthropic #OpenAI #JohnHultquist #Gemini #Claude #Mythos #OpenClaw #AdaLovelace #UKgovernment #aiartificialintelligence #business #cybercrime #cyberwar #hacking #technology