What is Prompt Injection Defense?
Techniques that stop untrusted content from hijacking model instructions: delimiting retrieved text, least-privilege tools, and confirmation before side effects. Treat all pasted or fetched content as data, never orders.
Related terms
System Prompt
The persistent instruction block prepended to every request that sets role, rules, and output format. It outranks user text in well-built stacks, which is why builders version it like code, not copy.
Function Calling
Letting a model request named tools with typed arguments instead of answering directly. The app executes the call and returns results, so the model can query data or trigger actions without inventing facts.
Grounding
Tying a model’s answer to retrieved sources instead of pure parametric memory. Grounded answers cite pages, which turns citation readiness into visible referral and brand presence.
Temperature
Sampling randomness for generation: low values make output focused and repeatable, high values make it varied and surprising. BYOB chat configs default near 0.5 while tool calls use 0.3 for steadier structured actions.
Thinking Budget
The token allowance reserved for internal model reasoning before answering. Larger budgets suit hard multi-step problems; BYOB prices thinking tokens separately so runaway reasoning shows up in metering.
Structured Output
Forcing model replies into a machine-readable shape such as JSON with fixed fields. Required wherever AI output feeds code, databases, or UI components, since free text breaks parsers the moment formatting drifts.