AI Spam Filters Tricked by "Text Salting

AI spam filters are reportedly being tricked by an old-school technique called "text salting." This method makes it harder for advanced AI to detect spam. What's interesting is that this highlights a vulnerability in current AI defenses. The Register also mentions other trending topics, like OpenAI admitting its GPT-5.6 occasionally deletes files, and a researcher poisoning an open-weight AI model for under $100. In security news, Russians are posing as Signal support for phishing attacks, and Microsoft patches failed to fix on-prem SharePoint, which is now under zero-day attack. The bottom line is that even sophisticated AI systems have weaknesses, and new threats continue to emerge in the digital landscape.