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.
Intel has officially suspended its long-running bug bounty program, eliminating cash payouts of up to $100,000 in favor of a zero-reward vulnerability disclosure system on the Intigriti platform. Launched in 2017, Intel's bounty initiative was once a cornerstone of its proactive security strategy, responsible for identifying nearly half of the company's addressed vulnerabilities in 2020 alone.