AI-Generated Malware.

Malicious code written, mutated, or assisted by large language models, lowering the skill bar for attackers and accelerating variant production. It belongs to the AI & ML Security category of cybersecurity.

Malicious code written, mutated, or assisted by large language models, lowering the skill bar for attackers and accelerating variant production.

AI-generated malware is created with the help of large language models or specialised code-generation tools, which produce loaders, droppers, infostealers, or polymorphic variants either on demand or as part of an automated pipeline. Attackers use both general-purpose models (often jailbroken) and underground 'malicious LLM' offerings advertised on cybercrime forums. Documented effects include faster development of phishing kits, on-the-fly obfuscation and string mutation to evade signature-based antivirus, and rapid translation of public proof-of-concept exploits into weaponised tooling. Defences focus on behaviour-based detection, EDR telemetry, sandboxing, model-provider abuse monitoring, and threat-intel sharing about prompts and artefacts associated with AI-generated tooling.

Defences for AI-Generated Malware typically combine technical controls and operational practices, as detailed in the full definition above.

Common alternative names include: LLM-written malware, AI-assisted malware.