Curated news, CVE analysis, and threat reports from the world's top cybersecurity sources.
Anthropic publicly disclosed that it identified and disrupted seven distinct China-based AI research laboratories operating large-scale Claude model distillation campaigns. These operations used thousands of automated API queries to systematically extract Claude's capabilities for creating unauthorized derivative models with reduced safety guardrails.
This article describes how the PaperCut AI swarm attack compromised 395 organizations and represents a new era in cybersecurity. Security researchers explain that this automated AI-powered attack marks a paradigm shift in how threat actors will conduct future operations.
This article reports that an attacker breached the email marketing platform Brevo through a SAML SSO vulnerability, accessing 138 customer accounts. The incident led to phishing attacks targeting cryptocurrency companies and their subscribers, highlighting supply-chain risks in email marketing services.
Anthropic reported that cybercriminals and state-sponsored threat actors are actively using Claude to automate exploitation and data theft operations across multiple victims. The AI is being employed at every stage of the attack chain, from generating convincing phishing lures to automating post-exploitation data identification and extraction.
Anthropic revealed that it disrupted a Russian state-sponsored hacking operation that was using Claude to rapidly rebuild malware tooling after their existing implants were detected and cleaned from victim networks. The attackers used the AI model to generate equivalent but code-signature-different malware variants, automating the rewriting of custom malware from scratch.
This article reports on how threat actors are abusing trusted AI platforms to host malicious content, poison search results, and trick users into installing malware. Huntress examines campaigns targeting AI users through weaponized Claude Artifacts, shared AI conversations, sponsored search results, and ClickFix-style lures.