ASCII Smuggling: Spammers Adopt Invisible Unicode to Evade Email Filters
Microsoft says email spammers are adopting ASCII smuggling
ASCII smuggling, a technique once used to hide malicious prompts in AI attacks, has been adopted by spammers to bypass email filters. By using invisible Unicode tag characters, spammers can obfuscate trigger words like 'funding' so that filters see 'fun' and 'ding' while humans see the original word. Microsoft detected a surge in such spam, with daily signature detections jumping from 21,000 to over 1.3 million in February 2026. The technique exploits the gap between human perception and text processing, posing a challenge for ML-based spam detection.
Because tag characters are invisible to humans but exist at the text-processing level, the same property that makes them useful for smuggling instructions into a model also makes them useful for obfuscating keywords before a detector evaluates them.
- CrimsonRain
Getting bombarded by "Paypal" <[email protected]> spam for months and Microsoft (outlook) can't handle that. This is beyond their level.
- gschizas
Why not sanitize and denormalize (or whatever it's called) the text before feeding it to the spam filter? Or the LLM prompt?
- k12sosse
As a [email protected] holder, unfiltered spam is less commonplace than
a) websites adding me to distribution/mailing lists without verifying I was the person to type it in,
b) people who either accidentally transform their address into mine, or forget part of their address.
These are from all parts of the globe. I have access to bank accounts in South America, Disney employee music royalty earnings tax disclosures and forms, European subscribers online platforms, AWS account recovery options, veterinarian records in Studio City, private school/PTA leadership website access in Mountain View.
I stopped trying to return unopened mail, nobody cared.
I don't do anything with any of this because I'm not a giant fool, but people, have your users verify their email addresses before you trust them.