Tuesday, August 11, 2026
LIVEThe Unrelenting Cyber Battle: Hacking Threats and the Imperative of Robust Data Protection///Navigating the Cyber Labyrinth: Bolstering Defenses Against Evolving Hacking Threats///The Dual Front War: Battling Hacking and Bolstering Data Protection in the Digital Age///The Ever-Evolving Cyber Threat Landscape: Navigating Hacking and Fortifying Data Protection///The Unseen Battle: Fortifying Data in an Age of Relentless Hacking///The Unseen War: Hacking's Relentless Advance and the Imperative of Data Protection///The Evolving Threat Landscape: Hacking, Data Protection, and the Imperative for Proactive Security///Navigating the Digital Minefield: Bolstering Data Protection in an Era of Relentless Hacking///The Dual Fronts of Digital Defense: Combating Hacking and Fortifying Data Protection///Hacking's New Frontier: Fortifying Data Protection in the Age of Advanced Cyber Threats///The Dual Front: Navigating Hacking Threats and Fortifying Data Protection in the Digital Age///Navigating the Digital Gauntlet: The Evolving Nexus of Hacking and Data Protection///The Unrelenting Cyber Battle: Hacking Threats and the Imperative of Robust Data Protection///Navigating the Cyber Labyrinth: Bolstering Defenses Against Evolving Hacking Threats///The Dual Front War: Battling Hacking and Bolstering Data Protection in the Digital Age///The Ever-Evolving Cyber Threat Landscape: Navigating Hacking and Fortifying Data Protection///The Unseen Battle: Fortifying Data in an Age of Relentless Hacking///The Unseen War: Hacking's Relentless Advance and the Imperative of Data Protection///The Evolving Threat Landscape: Hacking, Data Protection, and the Imperative for Proactive Security///Navigating the Digital Minefield: Bolstering Data Protection in an Era of Relentless Hacking///The Dual Fronts of Digital Defense: Combating Hacking and Fortifying Data Protection///Hacking's New Frontier: Fortifying Data Protection in the Age of Advanced Cyber Threats///The Dual Front: Navigating Hacking Threats and Fortifying Data Protection in the Digital Age///Navigating the Digital Gauntlet: The Evolving Nexus of Hacking and Data Protection///
Subscribe
Cyber Security
Independent · Digital
Thehackingpost
RSSSubscribe
CybersecurityAI-assisted

Simple Custom Font Rendering Can Poison ChatGPT, Claude, Gemini, and Other AI Systems

A recent examination has unveiled a vulnerability in AI web assistants, revealing a disparity between the visual rendering of a webpage and the textual data parsed by AI tools.

BBy Benjamin ScottMar 18, 20262 min read

A recent examination has unveiled a vulnerability in AI web assistants, revealing a disparity between the visual rendering of a webpage and the textual data parsed by AI tools.

By utilizing a custom font file and basic CSS, attackers can embed harmful instructions that bypass AI safety checks, which only perceive benign content.

The vulnerability was identified in December 2025, exploiting the difference between the Document Object Model (DOM) text of a webpage and its rendered visual output. AI assistants generally process the raw HTML, while browsers display the page using fonts, CSS, and glyph mappings, which can be manipulated by attackers.

LayerX demonstrated this vulnerability through a proof-of-concept webpage designed to appear as a fanfiction site for the video game Bioshock . The custom font used acted as a visual substitution cipher, displaying standard HTML text as indistinct gibberish while presenting a hidden payload as visible text, instructing users to execute malicious actions.

Various AI assistants, including ChatGPT, Claude, Copilot, Gemini, and others, were unable to detect the threat, often confirming the page's safety. The vulnerability does not require JavaScript, exploit kits, or browser flaws, as it resides in AI tools that equate DOM text with user-visible content.

By utilizing a custom font file and basic CSS, attackers can embed harmful instructions that bypass AI safety checks, which only perceive benign content.
Benjamin Scott · Thehackingpost

LayerX reported the findings to major AI vendors in December 2025. Responses varied significantly:

Vendor Response

Microsoft Accepted the report; requested a full 90-day remediation period Google Initially prioritized as high, later de-escalated and closed on Jan. 27, 2026 OpenAI Rejected as "out of scope" due to insufficient impact Anthropic Rejected as social engineering, considered out of scope xAI Rejected; directed to safety@x.ai Perplexity Classified as a known limitation, not a security vulnerability

Advertisement

Microsoft was the sole vendor to fully address the issue within the disclosure timeline.

The primary risk involves AI-assisted social engineering, where attackers leverage AI tools to endorse malicious pages. LayerX advises AI vendors to adopt dual-mode render-and-diff analysis, treat custom fonts as potential threats, scan for CSS-based content hiding methods, and avoid issuing definitive safety assessments without verifying the complete rendering context.

Based on reporting by Cyber Security News.

AI transparency. This article was produced with the assistance of artificial intelligence and published under human editorial oversight. AI systems can make mistakes. Read how we use AI (EU AI Act, Art. 50).
Related Stories
Cybersecurity

Why Fintech Banking Platforms Are Reducing Operational Costs by 40%

Paige Monroe · Mar 26, 2026
Cybersecurity

Dogecoin Price Prediction: SHIB Burns 172M Tokens as Pepeto Exchange Earns From Every Trade

Eric Wallace · Mar 26, 2026
Cybersecurity

Solana Price Prediction: SOL Holds $86 With Commodity Status as Pepeto Targets Bigger Returns

Zachary Burns · Mar 26, 2026
Cybersecurity

Open Directory Malware Campaign Uses Obfuscated VBS, PNG Loaders and RAT Payloads

Joseph Cain · Mar 26, 2026