A public-observatory booking generator is supposed to emit a UI script. Review finds an unexpected outbound “diagnostics” channel that nobody requested. Which security vector has the reviewer likely found?
Select an answer to reveal the explanation.
Short Explanation
If the script grows a secret phone line home that nobody ordered, review just caught malicious code generation. The point is finding unexpected dangerous content before it runs. No sample malware required — just refuse to treat surprise side effects as normal.
Full Explanation
Malicious code generation is the v1.1 vector where an LLM is manipulated into emitting code with harmful side effects, such as unexpected external command or channel calls. Testers discover this during review of generated scripts. The exam focuses on recognition and safe handling, never on sample payloads or generation prompts.