A Unicode homoglyph attack is a deception technique that uses characters from different writing systems that look alike, such as similar Latin and Cyrillic letters. It is used in domain names, user names and code to mislead readers, filters or review processes, sometimes to impersonate trusted identities or hide malicious text.
What This Attack Changes
Unicode homoglyph attack exploit the gap between what text looks like and what it actually contains. That makes them especially useful for misleading humans, bypassing weak string-based checks, and hiding hostile content inside names, links, or source text.
They are dangerous because review processes often depend on visual recognition. A malicious character choice can make a domain, username, code token, or label appear trusted while still being technically distinct, which is enough to support impersonation or stealth.
Where Homoglyph Attacks Appear
The most common targets are domain names, account names, file names, identifiers in code, and any user-facing text that passes through copy, paste, search, or moderation workflows. The attack is not limited to internationalised text, because the deception often depends on mixing scripts in a way the reader does not notice.
In practice, the risk grows where systems display identifiers in small fonts, truncate long values, normalise input inconsistently, or let users register visually similar names. A common example is a lookalike domain built from a Latin character replaced with a Cyrillic character that renders almost identically.
Why Detection Is Hard
Homoglyph abuse is difficult to catch with ordinary pattern matching because the malicious string may be valid Unicode. Defenders need more than blacklist thinking, since the same visual deception can be produced with many different character substitutions and combinations.
Reviewers also face a human factor problem: the closer a spoofed string is to a familiar brand, login name, or internal code reference, the more likely it is to be accepted at a glance. That is why this technique is often paired with phishing, brand impersonation, or malicious code obfuscation.
How Organisations Reduce Exposure
Resilience starts with strict handling of identifiers. Systems should normalise where appropriate, reject mixed-script registrations when there is no business need, and use display rules that make suspicious characters easier to spot. Where names matter operationally, the safest approach is to compare canonical forms rather than relying on visual inspection alone.
Readers looking for a broader view of domain impersonation and lookalike abuse can use the Phishing Guide, which covers lookalike domains and related deception patterns. For incident context, NHIMG’s The State of NHI & AI Agent Breach Report 2026 is useful when homoglyph-style deception is part of a broader credential or account compromise chain.
Risk and Threat Considerations
Unicode homoglyph attacks matter because they can turn a trusted-looking identifier into a control bypass, a phishing lure, or a review failure. The risk is highest where people or systems make decisions based on appearance rather than exact character identity, especially in identity, naming, and approval workflows.
Failure mechanism: An attacker introduces visually confusable characters so the target string passes human review, weak comparison logic, or superficial trust checks while remaining technically distinct.
Impact: The result can be credential theft, misdirected access, false approval, malicious code review evasion, or impersonation of a trusted brand, user, or service.
Standards & Framework Alignment
This section maps relevant standards and security frameworks to the operational risks and controls described in this guidance.
CIS Controls v8, OWASP ASVS and NIST SP 800-53 Rev 5 set the governance and control requirements practitioners need to meet.
| Framework | Control / Reference | Relevance |
|---|---|---|
| CIS Controls v8 | CIS-16 — Application Software Security | Homoglyph abuse exploits software and workflow weaknesses in input handling and review. |
| Recommendation — Apply CIS-16 to validate and canonicalise user-controlled text before it reaches trust decisions. | ||
| OWASP ASVS | V1 — Encoding and Sanitization | Confusable Unicode characters are an input-handling problem that affects how text is accepted and displayed. |
| Recommendation — Use V1 checks to normalise and validate text so confusable characters do not bypass review. | ||
| NIST SP 800-53 Rev 5 | SI-10 — Information Input Validation | Unicode homoglyph attacks rely on unsafe acceptance of deceptive input values. |
| IA-5 — Authenticator Management | When homoglyphs target usernames or account identifiers, safe identity handling depends on controlled credential and naming workflows. | |
| Recommendation — Enforce SI-10 to validate identifier content and reject unsafe character combinations. Apply IA-5 to manage identity-related values so lookalike strings do not weaken account controls. | ||
Practitioner Guidance
Why practitioners should care: This is not just a spelling trick, it is a trust problem. If your workflow accepts human judgment on names, domains, or code identifiers, homoglyphs can undermine that judgment without triggering a traditional security alert.
What to watch for: Mixed scripts, suspiciously similar names, and values that look normal only when rendered in a particular font or UI. The safer operational stance is to validate the exact characters, not just the visible shape of the text.
Practitioner takeaway: Treat visually similar text as untrusted until canonicalised, compared, and reviewed with character-level precision.
Related resources from NHI Mgmt Group
Deepen Your Knowledge
Free weekly newsletter
Subscribe to the NHI & AI Identity Journal
The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.
Bonus 33% off our NHI Course when you subscribe.
Reviewed and updated by the NHIMG editorial team on October 29, 2026.
NHI Mgmt Group — the #1 independent authority on Non-Human Identity, IAM, and Agentic AI security. nhimg.org