When Bannon and Sanders agree on AI regulation, pay attention

The news that Steve Bannon and Bernie Sanders agree on something is, by itself, remarkable. The fact that the thing is AI regulation should make every practitioner sit up.

I have spent years in program governance watching political consensus form slowly. Here it is forming fast. When voices from opposite poles of the U.S. spectrum converge on the need for rules, the window for self-regulation is closing.

Bannon and Sanders each have their reasons. Bannon worries about concentrated corporate power and cultural disruption. Sanders worries about worker displacement and algorithmic bias. Neither cares about my compliance headaches. But both are right that ungoverned AI creates real harm.

The practical takeaway, stop debating whether regulation will come. It will. The shape of it is what is still in play. If we want rules that work for real operations, we need to be in the room now.

That means writing clear model documentation. Building human-in-the-loop processes that are actually auditable. Fitting bias testing into deployment pipelines as a gate, not a checkbox. And explaining these choices in plain language to executives, not in techno-jargon.

The Bannon-Sanders moment tells me the public mood is shifting faster than most governance frameworks can handle. Companies that treat compliance as an afterthought will scramble later. The ones that treat it as a design constraint today will set the standard.

I am not interested in waiting for regulators to draw the lines. I prefer to draw my own, well within reason, and let the evidence speak. That is the only strategy that makes sense when the left and the right both have their eyes on the same target.