Raw vector
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:H/VI:H/VA:H/SC:N/SI:N/SA:N/E:P/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:XSummary
CVE-2026-33654 is a high-severity Code Injection (CWE-94) vulnerability in Nanobot Nanobot. Its CVSS base score is 8.9 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique Template Injection (T1221); ranked at the 40th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.
This vulnerability is AI-related — categorised as Enterprise AI Assistants; in the LLM/Generative AI Risks risk domain.
The strongest mitigations our analysis identified map to IA-2 (Identification and Authentication (Organizational Users)) and IA-3 (Device Identification and Authentication) — see the control section below for these in your framework.
Deeper analysis AI-assisted summary
Synthesised by an AI model from the NVD description and linked references — a reading aid, not an authoritative source.
CVE-2026-33654 is an indirect prompt injection vulnerability in the email channel processing module (nanobot/channels/email.py) of nanobot, a personal AI assistant. The flaw affects versions prior to 0.1.6 and has a CVSS v3.1 base score of 9.8 (AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H), with associated CWEs-94 (Code Injection), CWE-290 (Authentication Bypass), and CWE-1336 (Inequivalent Security Check in Implementation).
A remote, unauthenticated attacker can exploit this vulnerability by sending an email containing malicious prompts to the bot's monitored email address. The bot automatically polls, ingests, and processes the email content as highly trusted input, fully bypassing channel isolation. This enables execution of arbitrary LLM instructions and subsequently system tools without any interaction from the bot owner, resulting in a stealthy, zero-click attack with high confidentiality, integrity, and availability impacts.
The GitHub security advisory (GHSA-4gmr-2vc8-7qh3) states that version 0.1.6 patches the issue. Security practitioners should upgrade to this version to mitigate the vulnerability.
OWASP Top 10 for Web (2025)
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2026-16777
Vulnerability Data
nanobot is a personal AI assistant. Prior to version 0.1.6, an indirect prompt injection vulnerability exists in the email channel processing module (`nanobot/channels/email.py`), allowing a remote, unauthenticated attacker to execute arbitrary LLM instructions (and subsequently, system tools) without any interaction…
more
from the bot owner. By sending an email containing malicious prompts to the bot's monitored email address, the bot automatically polls, ingests, and processes the email content as highly trusted input, fully bypassing channel isolation and resulting in a stealthy, zero-click attack. Version 0.1.6 patches the issue.
- CWE(s)
AI Security AnalysisAI
- AI Category
- Enterprise AI Assistants
- Risk Domain
- LLM/Generative AI Risks
- OWASP Top 10 for LLMs 2025
- None mapped
- AI-specific weaknesses CR
- CWE-1427 — Indirect prompt injection via untrusted email input processed as trusted LLM instructions.
Mapped by Cyber Resilience · not in NVD. Poisoning and extraction cases are routed to MITRE ATLAS instead of a synthetic CWE.- Classification Reason
- Matched keywords: ai, llm, prompt injection
Related Threats
MITRE ATT&CK Enterprise Techniques
CVEs Like This One
Affected Assets
Mitigating Controls
Control response
—
—
- 3 hardening rules · 3 OS baselines
V1.3.2V1.3.7V1.3.10V6.4.3
Mitigating Controls (NIST 800-53 r5) AI
Proper unique identification and authentication of users directly stops spoofing-based bypass of authentication.
Device identification and authentication before connection prevents spoofing of devices to bypass auth.
Authentication of non-organizational users blocks external spoofing attempts against the scheme.
Authenticator management ensures credentials cannot be easily spoofed or reused to bypass authentication.
Developer testing and evaluation finds code paths that accept and execute externally influenced strings.
Input validation directly stops untrusted data from being used to construct executable code without neutralization.
Mitigating Controls (NIST CSF 2.0) AI
Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→CSF cross-walk (authority under review) — links open the control.
Protecting, conveying, and verifying identity assertions specifically prevents the spoofing that enables authentication bypass.
Requiring authentication of users/services/hardware directly counters spoofing-based bypass when strong methods are used.
PR.PS-06's SDLC practices directly target injection flaws via secure coding and testing (mostly), yet as a single broad outcome it leaves many code-generation specifics unaddressed (partial).
Proofing and binding identities reduces spoofing opportunities during enrollment but does not address runtime authentication implementation flaws.
PR.DS-10 protects runtime data confidentiality/integrity but has no bearing on neutralizing externally influenced input during code generation, so neither direction shows any preventive effect.
Mitigating Controls (ISO/IEC 27001:2022 Annex A) AI
Derived directly from the weakness types (CWEs) cited in the NVD entry via our AI-authored CWE→ISO cross-walk (authority under review) — links open the control.
Secure authentication control directly mitigates authentication bypass by spoofing.
Authentication information management directly addresses credential handling that prevents spoofing.
Security testing can detect spoofing vulnerabilities but does not prevent them by itself.
Access control policy reduces spoofing opportunities but does not prescribe authentication mechanisms.
Identity management supports unique identities but does not guarantee resistance to spoofing.
Access rights assignment limits exposure but does not enforce authentication strength.
Hardening callouts derived
Configuration rules from DISA STIG baselines that bear on weaknesses of the type cited by this CVE. Each rule is shown with the relationship its mapping actually records, against the CWE it was authored against. Derived via CVE→CWE over `controls_xwalks` (authoritative rows only; rows rated `none` are excluded).
Oracle Linux 8 (1 rule)
- V-248827 OL 8 must not have the rsh-server package installed. prevents CWE-290
RHEL 7 (1 rule)
- V-204442 The Red Hat Enterprise Linux operating system must not have the rsh-server package installed. prevents CWE-290
RHEL 8 (1 rule)
- V-230492 RHEL 8 must not have the rsh-server package installed. prevents CWE-290