President Joe Biden has unveiled a transformative executive order to oversee the booming artificial intelligence (AI) sector. Central to his mandate is the stipulation that AI should champion “civil rights” and “equity”, echoing the progressive tenets of diversity, equity, and inclusion (DEI).
Under President Biden’s new Executive Order on AI, the federal government will help develop standards to watermark and clearly label AI-generated content – so people can identify what’s real and what’s not.
— The White House (@WhiteHouse) October 30, 2023
To combat potential biases in AI, The White House is prepping to dispatch guidelines to various entities like landlords, federal contractors, and benefits programs. This comes hand-in-hand with a rigorous enforcement plan, as the order spells out a collaboration between the Department of Justice and Federal civil rights offices. Their mission? To outline the gold standards in scrutinizing and acting against civil rights breaches related to AI.
A cornerstone of this directive targets the justice system’s AI applications. The administration aims to roll out protocols for AI involvement in processes such as sentencing, parole, risk assessments, and even predictive policing.
But beneath the surface lies a brewing debate: Could a hyper-focused AI, when fed with datasets like criminal records or credit scores, inadvertently perpetuate inequalities? This conundrum finds its roots in machine learning fairness, a novel discipline bridging critical race theory and tech. The crux is ensuring AI decisions do not inadvertently defy U.S. civil rights, meaning some AI systems might need to suppress their predictive precision.
Diving deeper, the executive order envisions an AI landscape where authenticity matters. Companies crafting AI-generated content will be mandated to embed watermarks, demarcating AI from human-produced content.
Safety remains paramount. With the National Institute of Standards and Technology setting the benchmark for AI safety tests, the Department of Homeland Security is on deck to launch an “AI Safety and Security Board”, guaranteeing AI in critical sectors adheres to safety norms.
Yet, the inclusion of DHS in AI oversight rings alarm bells, especially amidst concerns of free speech advocates. Given DHS’s history in curbing “disinformation” and its attempts at “deradicalizing” certain demographics, many are wary.
On the security frontier, there’s an inter-agency effort between the Departments of Energy and Homeland Security to confront potential AI threats, spanning from cyber to biological risks.
Employee safeguarding also makes the cut. While the order hints at strategies to protect jobs, specifics are sparse. A contradictory angle emerges with an immigration aspect; the White House envisions revising visa policies to draw AI talent from overseas.
Privacy remains a pivotal element. The administration plans to scrutinize how AI data is procured, especially by federal agencies. The aim? To back ventures that train AI while upholding data privacy.
In essence, while Biden’s order paints a visionary future for AI, it raises pressing questions on privacy, freedom, and the role of big government in tech.