Policy Dive: AI in the First Week of Trump

Trump’s view of AI development has changed even from his first term, when his administration wrote that AI development must happen “while maintaining the safety, security, civil liberties, privacy, and confidentiality protections we all expect.” His new Executive Order, combined with virulent takedowns of any indication of equity or anti-bias efforts across the government, indicates a preference for policies that will prioritize speed over safety, acquiesce to the desires of Big Tech at the expense of the environment and civil rights, and strip the government of much-needed guidance in how to implement AI safely. For the moment, the EO appears to be more of a blunt instrument than a thoughtful progression, creating more chaos than opportunity. It remains to be seen how the AI Action Plan and OMB guidance will address this vacuum of leadership.

Sir Demis Hassabis, the Nobel-Prize winning chief of Google DeepMind, warned in an interview about full-speed, guardrail-free AI development, saying “There’s much more at stake here than just companies or products [...] it’s the future of humanity, the human condition and where we want to go as a society.” Without measures in place, AI models have the potential to codify racism, sexism and homophobia, and income inequality, exacerbating existing cultural, class, and global divisions. States will likely step up to implement their own guidelines, resulting in an American patchwork of regulations that will create concerns (and lawsuits) around data usage and privacy, civil rights, and discrimination across state lines.