Intel has the same (or higher level) of access to AI tools as any researcher. Intel could be using these AI toolsets to design their chips, software/drivers to harden them, to the point that the number of bug submissions outside that scope would be minimal.
The only other challenge would be reviewing submissions, and, again, there are AI toolsets for just this, (with human oversight, of course.)
Long term, I think fewer eyes on your bits means you rely more on security through obscurity, which never works out in the end.
Good luck, Intel!