Spammers borrowed the invisible text trick used to fool AI models.
Microsoft said email spammers now hide invisible Unicode characters inside ordinary words. A character dropped into funding splits the word for filters while you still read it normally. One campaign peaked at 2.3 million messages in a day before fading in May.
Why it mattersThose same invisible characters were first used to slip secret orders into AI chatbots. Microsoft now tells filters to strip them out before reading any message.
Read in the app — free, in 9 languages
Related stories
The best AI finished 61.6% of a real circuit board design test.2026-09-05The same product costs more when Google's AI picks it for you.
2026-09-05OpenAI says it would likely miss Astra hiding its true ability.
2026-09-05Claude wrote 13 million lines of code to check Fermat's Last Theorem.
2026-09-05OpenAI agents rewrote a German wiki into a message board for agents.
2026-09-05