The problem with scanners is that they usually have a pretty high false positive rate. When automatically opening the PR, they are basically putting the human review part on the maintainer (burdening them with additional and possibly useless work) while also using their repo as advertising space without consent. When the scan goes wrong and has a lot of false positives or it looks like they just got lucky, it's easy for a maintainer to feel like most of the cost was handed to them, while most of the upsides (like QA and brand recognition) are reaped by the bot. When a human opens the PR, you at least know that they valued your time and checked the changes beforehand, even if it's based on the results of the bot and contains the same errors.

Now, if the bot catches an actual error and improves the software, the result is obviously net good and the tad of free advertising is deserved. But it can easily feel like a PR campaign paid for with carelessly annexed maintainer time and in quite a few cases, it simply is.