Lack of AI governance is a real threat
August 31, 2026
In his latest Quick Take, Ian Bremmer examines why artificial intelligence has become one of the most politically charged issues in America — and why almost nobody is actually governing it.
AI and the data centers powering it are now driving a significant share of US economic growth, but public skepticism is rising fast. Some of that resistance is straightforward nimbyism: concerns about zoning, energy costs, and eyesores, while some reflects a deeper distrust of the tech CEOs and hyperscalers benefiting most from the buildout. Politics plays a role too, as critics tie Trump to the billionaires who backed him, and disinformation from actors like Chinese bot farms further muddies the debate.
Ian remains a longtime proponent of AI's potential, but argues that the near-total absence of governance is becoming untenable, pointing to a recent incident in which OpenAI agents escaped a sandbox environment and infiltrated Hugging Face, illustrating what's known as the "paperclip problem," where an AI single-mindedly pursuing one goal can produce dangerous, unintended consequences.
Even more striking, over a thousand employees at frontier AI companies — along with figures like Bill Gates have urged the US government to slow development down. "I'm increasingly convinced that they are right, but that doesn't mean it's going to happen," Bremmer says. With political leaders lacking expertise, companies unwilling to self-regulate, and China showing no signs of pausing its own AI race, Bremmer warns that the risks of unchecked AI development are growing faster than our ability to understand them.
Keep up with what’s going on around the world - and why it matters.