More importantly a malicious actor narrowly targeting a specific project doesn't need an LLM to achieve the same result. An LLM might make it a bit simpler to scale this up to many projects but on an individual project level posing as a minor contributor for a while is quite feasible manually too.
此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。