2026-09-10 · Fuente：Ars Technica Tech Policy
A once-overlooked block of Unicode invisible to humans, ASCII smuggling was a favored way to attack AI systems — and is now being embraced by spammers.
ASCII smuggling — a technique built on a block of Unicode characters that are invisible to human readers — has found a new set of users. According to reporting published by Ars Technica's security and tech policy desk in September 2026, the method, once a popular means of attacking AI systems, has now been embraced by spammers, and the once-overlooked Unicode behind it is seeing ever wider use.
The technique takes advantage of a portion of the Unicode standard that was once overlooked. The characters in that block do not render visibly, meaning text containing them can go unseen by a person reading it. That invisibility is what made the block attractive as a component of attacks aimed at AI systems, where it enjoyed a period of genuine popularity before attention moved elsewhere.
The notable change is who is deploying it now. Rather than remaining a niche instrument for attacks on AI systems, ASCII smuggling has been picked up by spammers, giving the technique a second life well beyond the context in which it first became popular. The invisible Unicode at its core, long overlooked, is now described as being put to ever wider use across that expanding base of users.
The migration matters for anyone working on the defensive side of content security. A trick that first drew attention as an AI attack vector has crossed over into everyday abuse, with the obscure characters that power it now circulating more widely. An overlooked corner of the Unicode standard has thus moved into broader circulation — a reminder that techniques developed against AI systems do not necessarily stay confined to that arena.
Etiquetas: ASCII Smuggling · Unicode · AI Security · Spam
萬安算交所 AIXX · AIXX · Exchange verificado · Ir al mercado