/

CVE-2026-92913

Severity:

HIGH RISK

(7.4)

Weak PRNG (CWE-330) in AVideo Account Activation Codes

CVE-2026-92913

Deferred

Vulnerability type

Category:

Use of Insufficiently Random Values

CWE:

CWE-330

Impact:

Account Takeover and Credential Disclosure

Severity:

HIGH RISK

(7.4)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

AVideo generates account activation and login pairing codes with predictable values derived primarily from timestamps rather than a cryptographically secure random source. Because the code-generation endpoint is unauthenticated, a remote attacker may predict a valid, unexpired code and exchange it for account information and a long-lived credential, potentially enabling account takeover.

Affected context

The vulnerability occurs in AVideo deployments that generate activation or login pairing codes through the affected getRandomCode() function. Risk is increased when the unauthenticated login-code API is reachable and generated codes remain valid for their configured lifetime, allowing an attacker to estimate generation timing and attempt code guesses.

Exploit preconditions

  • Attack surface: Network-accessible AVideo login-code API endpoints
  • Required access: No authentication is required to request or analyze codes
  • Trigger point: Generation and subsequent redemption of an account activation or login pairing code
  • Why exploitation is possible: Codes are derived largely from predictable timestamp components, provide a relatively small search space, and can be tested during their validity window
  • Additional condition: A valid code must be guessed or otherwise recovered before it expires

Why this happens

The code-generation routine relies on uniqid() and a non-cryptographic rand() value instead of a cryptographically secure random number generator. Timestamp-derived values expose information about when a code was created, while the small padding component does not provide sufficient unpredictability. This maps to CWE-330: Use of Insufficiently Random Values, because security-sensitive tokens can be predicted or enumerated.

Potential impact

  • An attacker may obtain a target account's email address
  • An attacker could acquire a credential accepted in place of the account password
  • User accounts may be taken over for the credential's validity period
  • Attackers may access or modify account-associated data
  • Compromised accounts could be used for follow-on abuse or lateral activity

General mitigation guidance

Mitigation generally centers on replacing predictable token generation with cryptographically secure randomness and reducing the value of any intercepted token.

Common defensive patterns include:

  • Generate security-sensitive codes with a CSPRNG and sufficient entropy
  • Use short, single-use, purpose-bound tokens with server-side expiration
  • Bind pairing or activation tokens to the intended account and transaction
  • Avoid exposing timing or generation metadata through unauthenticated APIs
  • Require authentication or additional verification before revealing account-linked credentials
  • Apply rate limits, monitoring, and lockouts to code generation and redemption
  • Limit the privileges and lifetime of any credential derived from a temporary code

How teams detect issues like this

Organizations can identify similar weaknesses through:

  • Static analysis rules that flag uniqid(), predictable timestamps, or ordinary PRNGs in authentication and token-generation paths
  • Code review checks for security tokens assembled from timestamps or low-entropy padding
  • Tests that measure token entropy, uniqueness, predictability, and resistance to generation-time estimation
  • API review for unauthenticated token or code issuance and redemption endpoints
  • Policy enforcement requiring approved CSPRNG APIs for authentication material
  • Monitoring for high-volume code requests, repeated redemption failures, and successful redemption from unusual sources

Automated security review tools may help identify weak randomness patterns, but findings should be validated against token lifetime and server-side redemption behavior.

Sources

SourceLink
github.comgithub.com/WWBN/AVideo/security/advisories/GHSA-v65f-hc7x-wj62
vulncheck.comvulncheck.com/advisories/avideo-weak-prng-activation-code-authentication-bypass

Frequently asked

questions

Yes. The relevant code-generation API is described as unauthenticated and reachable over the network. Successful exploitation still requires guessing a valid code within its expiration period.

No existing account credentials are generally required to request codes or attempt redemption. The attacker does need a reachable target endpoint and a valid code guess.

Replace predictable, timestamp-derived codes with high-entropy values generated by a cryptographically secure random source. Make codes short-lived, single-use, account-bound, and subject to rate limits.

Treat activation and pairing codes as credentials, enforce approved CSPRNG usage in code review and CI policies, avoid unauthenticated token oracles, and monitor both issuance and redemption activity.

Find out what's already
exploitable in your codebase.