Are the Midterms Ready for a More Powerful AI?

We are watching electoral history unfold in America: The first non-human 2026 midterms. In the past weeks, we have seen evidence of many AI’s “misalignments” that have rapidly growing power to sway voters in ways that we do not fully understand. Did we miss the signs? How far can we trust AI leading up to November 3rd? What can we do to protect our votes? Abdiaziz Ahmed (“Abdi”) will open his presentation with the surprising findings from his six-month study that revealed chatbots who were already not following orders, had the ability to create deceitful images, audio and video. Abdi will explain in conversation what all of this means for voters, and what could change even days before November 3rd. Come prepared for uncertainty!

Speaker Abdiaziz Ahmed is the technology policy strategist at the Brennan Center Elections and Government Program, where he focuses on artificial intelligence safety, election security, and the impact of emerging technologies on public policy. He previously served as policy manager at Amazon and deputy policy director of the Democratic Attorneys General Association. He holds a degree from Carleton University.