American Exceptionalism in AI: Safety and Progress are Not Zero-Sum
Europe has already sealed its fate as an AI loser. The hasty reaction towards heavy regulation with the EU AI Act will not only discourage AI development but also heavily restrict European businesses' ability to benefit from AI. The impact of an AI dead zone in Europe is evident with Apple’s recent announcement to indefinitely defer the release of its AI in Europe.
Other nations, including Japan, the UK, and India, are considering similar legislation as a knee-jerk reaction to AI's potential. These regulations seek to limit AI capabilities in a misguided attempt to ensure public safety. However, these restrictions will only starve citizens and businesses of beneficial AI. More troublingly, they create a false sense of security against real AI threats on the horizon.
A Crucial Opportunity for American Leadership
In a time of declining US technological leadership, AI presents a crucial opportunity for American leadership. It's imperative that regulators and politicians adopt a forward-thinking approach to AI regulation. There will always be threats from AI, and regulation is a losing game of whack-a-mole.
Existing open-source models are already powerful enough to cause significant harm. By the time any regulatory regime and enforcement come into full effect, these models will have only improved. Even if severe regulations reduce the availability of powerful open-source models, training small, domain-specific models with high accuracy remains hobbyist-level simple. Regulators will always move slowly, while cybercriminals move fast, making this a losing battle.Sophisticated cybercriminals are already using AI to conduct spear phishing attacks at scale with little cost. This is just the beginning of smart, well-executed impersonation attacks that are incredibly real, deeply personalized, and increasingly difficult to spot.
Even if global AI restrictions make it difficult for criminals to use malicious AI, state-sponsored adversaries pose a much larger danger.Russia or China will likely seize the opportunity of a regulation-stifled West to wage a cold war with advanced AI. North Korea could execute effective cyberattacks sponsored by AI from China or Russia, given China’s historic backing of North Korean hacking groups and the increasingly close North Korea-Russia relationship.
Achieving AI Safety Through Deeper Regulation of Identity and Provenance
AI is dangerous because it can pretend to be anyone and generate images and videos that seem real. Trying to stay ahead of AI with detection mechanisms like reCAPTCHA or other AI models designed to detect AI-generated content is costly and ineffective. The real solution is creating universal systems of verification for digital identity and content provenance.
Governments should collaborate with industry to push forward standards and regulations that make this kind of verification cheap, scalable, and universally enforced. Some early steps are being taken to create government-verified digital IDs, but this work needs drastic expansion in partnership with identity providers focused on portable digital identity (like Footprint).
Regulations should be placed on software vendors that pose a distinct risk of impersonation attacks, such as email providers, SMS and voice services, and social media platforms, to use verified digital identification as the only form of access. Qualcomm is already pioneering hardware-level video and image provenance, guaranteeing an image was taken via a camera. Governments should push the handful of companies responsible for the firmware and operating systems of most devices to guarantee the availability of this technology. The government also has a key role in educating the public on a verification system and the importance of paying attention to it.
Seizing the AI Advantage
AI is more than just a technological breakthrough; it's a strategic asset that can cement America's leadership on the global stage. By driving economic growth, AI can spur innovation, create new industries, and generate high-paying jobs. Leading in AI will enable the US to set global economic standards, keep American businesses competitive, and attract top talent. In terms of national security, AI empowers the development of advanced defense systems and strengthens cybersecurity. Maintaining a strategic edge in AI allows the US to deter threats and assert its dominance in global security, safeguarding its interests and allies. On a softer note, by setting ethical standards, the US can influence international norms and foster global cooperation. This enhances America's reputation as a responsible leader, promotes American values worldwide, and strengthens diplomatic ties.
The Path Forward for American AI Leadership
The race for AI dominance is heating up, and the stakes have never been higher. As Europe stifles its own progress with heavy-handed regulation, the US stands at a crossroads. Embracing AI with a balanced approach to regulation—focusing on identity and provenance—will ensure safety without hindering progress. The US must seize this opportunity to lead in AI, driving innovation, enhancing security, and shaping global standards. The future of American exceptionalism depends on it.