PulseAugur
EN
LIVE 19:09:09

AI prompt injection technique now used by spammers to evade email filters

By PulseAugur Editorial · [3 sources] ·

A technique known as ASCII smuggling, initially developed to make AI prompt injection attacks more covert, is now being exploited by spammers to bypass email filters. This method uses invisible Unicode characters that are readable by computers but undetectable to humans. Microsoft Defender for Office 365 observed a dramatic surge in spam messages employing ASCII smuggling, with daily detections escalating from thousands to millions, indicating a significant shift in how malicious actors are attempting to deliver unwanted content. AI

IMPACT Spam campaigns are evolving to bypass AI-powered defenses, necessitating new detection strategies.

RANK_REASON A security technique developed for AI attacks is being repurposed for spam, with a significant increase in detection rates.

Read on Mastodon — mastodon.social →

AI-generated summary · Google Gemini · from 3 sources. How we write summaries →

AI prompt injection technique now used by spammers to evade email filters

How we ranked this

Signal score
19 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
A security technique developed for AI attacks is being repurposed for spam, with a significant increase in detection rates.
Source corroboration
3 independent sources
Multiple independent publishers reporting the same story raises confidence that it's real and newsworthy.
Topics
other
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
Breaking (< 6h)
Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.
Coverage growth since scoring
+1 source(s) since last score
New sources have picked up this story since our last re-score. Score will update on the next scoring pass.

Full methodology in our editorial standards.

COVERAGE [3]

  1. Ars Technica — AI TIER_1 English(EN) · Dan Goodin ·

    Once popular for attacking AI, ASCII smuggling is embraced by spammers

    A once-overlooked block of unicode that's invisible to humans is gaining ever wider use.

  2. Mastodon — mastodon.social TIER_1 English(EN) · [email protected] ·

    📰 Once popular for attacking AI, ASCII smuggling is embraced by spammers A once-overlooked block of unicode that's invisible to humans is gaining ever wider use

    📰 Once popular for attacking AI, ASCII smuggling is embraced by spammers A once-overlooked block of unicode that's invisible to humans is gaining ever wider use. 📰 Source: Ars Technica 🔗 Link: https://arstechnica.com/security/2026/09/once-popular-for-attacking-ai-ascii-smuggling-…

    LINKS arstechnica.com/…/once-popular-for-attack…
  3. Mastodon — mastodon.social TIER_1 English(EN) · [email protected] ·

    Once popular for attacking AI, ASCII smuggling is embraced by spammers https://arstechnica.com/security/2026/09/once-popular-for-attacking-ai-ascii-smuggling-is

    Once popular for attacking AI, ASCII smuggling is embraced by spammers https://arstechnica.com/security/2026/09/once-popular-for-attacking-ai-ascii-smuggling-is-embraced-by-spammers/ # Security # AI # Spam

    LINKS arstechnica.com/…/once-popular-for-attack…

RELATED ENTITIES

RELATED TOPICS