Lin Hsin Hsin Artificial Intelligence Center

TAXONOMY of Adversarial AI
















With increasing sophistication of multifaceted attacks on large language models (LLMs) across industries worldwide marks a dangerous uptrend in cyber threats targeting AI systems.

These attacks exploit the expanding capabilities of LLMs, manipulating them through adversarial inputs, prompt injections, model inversion, and data poisoning, often with little trace. As LLMs become increasingly integrated into sectors such as finance, healthcare, defense and education, the risks of misuse, data breaches, and systemic disruption escalate dramatically. Therefore, cybersecurity in AI must be proactively scrutinized and fortified ahead of time, with robust frameworks, continuous monitoring, and ethical safeguards to ensure resilience against evolving threats and preserve trust in AI-powered technologies.

Arm with in-depth research and extensive experiences in both Artificial Intelligence and Cyber Security, the founder of the Lin Hsin Hsin Intelligence Center has meticulously authored the platform and LLM agnostic TAXONOMY of Adversarial AI by category and sub categories. This framework will be kept up-to-date with the evolving threat landscape.