At the start of what will surely be a banner year for artificial intelligence (AI) systems, McSilver Executive Director Rose Pierre-Louis and U.S. Representative Adriano Espaillat co-authored an op-ed in the New York Daily News reflecting on challenges already encountered in the quest for more ethical and equitable AI systems and offering novel legislative solutions that could help shape the future of automated systems nationally and globally. The timely editorial points to bills like H.R.649 as tools to foster regional cooperation around policy encouraging a more just AI industry. Read the piece at the Daily News or see the full text quoted below:
Artificial intelligence (AI) is transforming our world in unprecedented ways. It can enhance our productivity, creativity, and well-being. It may help us solve some of the most pressing challenges of our time, such as climate change, poverty, and health care.
But machines can also discriminate.
They can, and often do, reflect and amplify the biases of their creators and data sources. A study of IBM, Microsoft, and Amazon’s AI systems’ abilities to guess the gender of a face found them to be excellent in identifying light-skinned men, resulting in an error rate of less than 1%. Yet for darker-skinned women the machines had significant trouble, misclassifying them at a rate of up to 35%.
These biases have real-world impacts. Facial recognition software, which relies on AI, can wrongly identify people of color and women, leading to false arrests. Without guardrails, AI can and will negatively affect underrepresented communities and communities of color.
Yet, the world now finds itself undergoing an ongoing “AI revolution.” In the past four years alone, the percentage of enterprises employing AI grew 270%. Indeed, governments and businesses have been increasingly relying on AI-based learning algorithms for daily functions such as credit scoring, hiring, health care, and education, all of which have a vast potential to discriminate.
Ultimately, AI algorithms are only as aware of unconscious biases as the humans who code them. Without a concerted effort to avoid bias in AI programming, these industry-specific AI algorithms are doomed to repeat the mistakes of AI facial recognition software.
Recognizing this, the Biden administration issued a white paper in October 2022 to address these concerns, called the Blueprint for an AI Bill of Rights, and, more recently, an executive order on new standards for AI safety and security. These are intended to support the development of policies and practices that protect civil rights and promote democratic values in the building, deployment, and governance of trustworthy AI.
Despite this, Congress has yet to act on regulating automated systems.
The internet also does not recognize borders, so comprehensive legislation that ensures AI machine learning models’ inputs and algorithms account for diversity and fairness will only be effective if our global allies adopt similar models. The U.S. needs to implement a regional AI strategy in the Western Hemisphere that protects the rights and interests of all people.
That is why we are urging Congress to support H.Res. 649, introduced by Rep. Espaillat, which calls for such a strategy. This resolution will ensure that machine intelligence is a tool used for the public good, promoting ethical and inclusive systems. It would also fortify America’s position as the leading force in the international AI race through three key provisions:
First, it will develop and implement a safe and responsible Regional AI strategy in the Americas guided by principles that will protect and advance civil rights, civil liberties, privacy, equal opportunities, and access to critical resources and services.
Second, it will leverage government funding from U.S. economic and international development agencies to invest in research and development (R&D), AI education, training, and infrastructure throughout the Western Hemisphere.
Third, it will increase the government’s cooperation and dialogue with international organizations, such as the Organization of American States (OAS) and the Inter-American Development Bank (IDB), to share resources and coordinate in the implementation of ethical rules for AI governance, investment, and development.
While we lay the groundwork for responsible technological development, new AI-driven threats are already here, including threats to democracy itself.
Before the 2024 presidential election, Rep. Espaillat also authored H.R. 4611 — the Candidate Voice Fraud Prohibition Act. This legislation seeks to protect working-class families from misinformation generated by automated systems and prohibit using maliciously unreal, albeit convincing, fabricated voice audio (“deepfakes”) of a candidate’s voice in campaign ads.
Continued inaction by the U.S. and our democratic allies to protect voting rights in the face of these new AI threats is not an option and is just one of several new issues that AI presents for the Hemisphere.
The U.S. has a unique but short window of opportunity to shape the direction of AI regulations in the Western Hemisphere in a way that ensures justice and fairness in AI. H.Res. 649 would help move the Americas to the forefront of developing these regulations and policies that will govern our future world. We must make AI work for us, not against us — before it is too late.
Espaillat represents New York’s 13th congressional district covering uptown Manhattan and parts of the Bronx. Pierre-Louis is executive director of the NYU McSilver Institute for Poverty Policy and Research.
