Skip to content
Integrating OpenAI into Your Application: A Security Perspective
hackerone.com | blog | #ai-security | #prompt-injection | #xss | #llm | #application-security | #openai | #rate-limiting
Summary
When integrating OpenAI models, treat output as untrusted: rate-limit API use, cap output length, sanitize against XSS and injection, and keep sensitive data out of prompts to limit jailbreak risk.
- Published
- Collected