Glossary

Defensive Prompting

Defensive Prompting embeds protective mechanisms directly into AI prompts to safeguard large language models (LLMs) and agents against adversarial inputs, prompt injections, and manipulations that could lead to unsafe or unintended behaviors. It ensures reliable, policy-compliant responses in high-stakes automation environments by prioritizing robustness over raw performance.

Defensive Prompting embeds protective mechanisms directly into AI prompts to safeguard large language models (LLMs) and agents against adversarial inputs, prompt injections, and manipulations that could lead to unsafe or unintended behaviors. It ensures reliable, policy-compliant responses in high-stakes automation environments by prioritizing robustness over raw performance.