Secure Go-LLM apps: Learn to block prompt injection, secure tool calls via least-privilege, and stop RAG data poisoning with robust patterns

In as more Go applications use LLMs in one way or another, security is more important than ever.

In this talk, we’ll explore some design patterns to keep your application secured against a range of attacks, such as:

• Prompt injection — How injected instructions in user input or retrieved documents can hijack model behavior.
• Securing tool calls — Apply least-privilege principles to exposed functions, prevent command injection from model output, and enforce authorization checks before execution.
• RAG pipeline security — Guard against data poisoning and indirect injection via ingested documents. See how a malicious document in the vector DB can manipulate retrieval results and model responses.

Join us to explore this critical area and learn together how to write safer applications.