Should moderators run automated bots under their accounts?

The most recent Stack Overflow election brought up some great new moderators, including Andy, who has for the past ~3 years been running a comment flag bot to automatically flag comments for removal. Since the end of the election, Andy graciously turned off the bot until we determined how best to proceed with it possibly running under his account.

During the time the bot was running, it flagged approximately 109,772 comments. Of these:

• 109,494 were marked helpful, and
• 278 were declined

That means the bot was accurate ~99% of the time.

Now that Andy is a moderator on the site, there are a few things that need to be discussed and decided before turning it back on:

1. Should it run under his main account, with mod privileges?
2. Should it run under another account, one which is controlled by Andy but without mod privileges attached?

Before chiming in on these, here are some things to think about:

• If the bot is run under his main account, then it would have the same privileges that he has as a moderator. This means there would no longer be a limit to the number of comments that could be deleted per day - the normal 100 flags per day limit would not be in effect. The bot could continue to delete comments without restrictions.
• If we do decide to let the comment-flagging bot run under a moderator account, it also means there will no longer be a pair of human eyes on the comment to make the final decision about deletion. A flag raised by the bot will auto-delete the comment and not send it to the mod queue for additional review.
• Like any other decision made by a moderator, everything is logged and tracked in the database, so if the bot deletes thousands of comment we have record of everything it does.
• Finally, if the bot is run under a moderator account, then any action it takes would fall under the moderator agreement that Andy signed post-election, meaning his main account would suffer any consequences if we determined that the bot was behaving badly.

With all that in mind, do we think it's appropriate for a moderator to run a bot under their main account to flag/delete comments unilaterally? Or should it be run separately from an account without moderator privileges?