OpenAI has discovered five influence operation (IO) campaigns utilizing its AI tools, ChatGPT and DALL-E. The IO campaigns, originating from Russia, China, Iran and Israel, created deceptive social media content but had limited public influence. OpenAI shared its threat intelligence with others in the industry and shut down the related accounts. It utilizes the data from such investigations to enhance its safety systems against future threats.

Mandiant warns of attacks on newly-disclosed Ivanti remote takeover threat
Google’s Mandiant team has issued an alert about a remote code execution flaw in the Ivanti Connect Secure VPN platform. The vulnerability, designated CVE-2025-22457, is