WASHINGTON, July 5 — The global chessboard for artificial intelligence regulation remains fragmented three years after the European Union’s landmark AI Act was proposed, with the United States and international bodies still maneuvering for position. Strategically, the EU’s AI Act, proposed in April 2021 and adopted in March 2024, set a high-water mark by categorizing AI applications by risk level and banning high-risk uses like social scoring. That move gave Brussels an early lead in setting rules for a technology that now touches everything from hiring to law enforcement.
What this signals is a deepening transatlantic divide in regulatory philosophy. The U.S. White House issued an Executive Order on Safe, Secure, and Trustworthy AI in October 2023, requiring safety assessments and transparency from AI developers.
But that order, unlike the EU’s legislation, lacks the force of a comprehensive statute. The broader stakes became clear in November 2023, when the U.S. and 28 other countries signed the Bletchley Declaration on AI safety. That non-binding agreement signaled a shared concern but no unified enforcement mechanism.
The calculus for American policymakers: move fast enough to keep pace with Beijing while avoiding rules that could stifle domestic innovation. The United Nations has also entered the fray.
Secretary-General António Guterres called for a global AI watchdog in 2023, though no such body has been established. The UN’s role remains largely consultative, a forum for debate rather than rulemaking. Ethical concerns continue to drive the regulatory push.
Algorithmic bias has been a persistent flashpoint, particularly in facial recognition systems. Amazon’s Rekognition, released in 2016, and IBM’s similar tools drew criticism for racial bias. The AI Now Institute at New York University, founded in 2017, has studied these impacts and pressed for stronger oversight.
The future scenarios debated by figures like Elon Musk and the late Stephen Hawking range from beneficial augmentation to existential risks. That spectrum of possibility — from tools that enhance human capability to systems that could pose catastrophic threats — shapes the political calculus in Washington and Brussels alike.
For now, the regulatory landscape remains a patchwork. The EU has law. The U.S. has executive action.
The international community has declarations. What this signals is a race not just to regulate, but to define the terms of engagement for a technology that is outpacing governance at every turn.
