The site's wording quantifies over every prime $p$ and fails at the smallest one: no prime is smaller than $2$, so $p=2$ has no prime primitive root $q<p$. A comment of 17 August 2025 by Woett in the site's discussion thread makes this remark, that $p>2$ is required, and the formal-conjectures statement assumes $p\ne2$. The change inserts "$>2$" after "every prime $p$"; nothing else changes. The defect is already in the poser's text: Erdős asks the question for every prime with no restriction, in [Er61e], p. 11, and in [Er65b], printed p. 233: "As far as I know it is not even known whether to every $p$ there is a prime $q<p$ which is a primitive root of $p$." The problem's standing judges the corrected Statement.