Offices would focus on bias, discrimination, and other harms resulting from algorithms

Bill Text (PDF)

Washington (December 12, 2023) – Senator Edward J. Markey (D-Mass.), a member of the Senate Commerce, Science, and Transportation Committee, today introduced the Eliminating Bias in Algorithmic Systems (BIAS) Act to ensure that every federal agency that uses, funds, or oversees artificial intelligence (AI) has an office of civil rights focused on combatting AI bias and discrimination, among other harms. The legislation would also require every civil rights office to report their efforts to Congress and provide recommendations for congressional action. 

Many federal agencies lack civil rights offices whose principal mission is to protect vulnerable communities across the United States. Currently, there are only 30 civil rights offices within the federal government, yet many are not required to secure staff with the expertise needed to advise the agency on algorithmic bias and discrimination. 

“From housing to health care to national security, algorithms are making consequential decisions, diagnoses, recommendations, and predictions that can significantly alter our lives,” said Senator Markey. “As AI supercharges these algorithms, the federal government must protect the marginalized communities that have already been facing the greatest consequences from Big Tech’s reckless actions. My Eliminating BIAS Act will ensure that the government has the proper tools, resources, and personnel to protect these communities and mitigate AI’s dangerous effects, while providing Congress with critical information to address algorithmic harms.”

Algorithms are already discriminating against disadvantaged communities in a variety of contexts. For example, companies have used algorithms to exclude users of certain gender identities from viewing job advertisements, block advertisements for clothing made for people with disabilities, and charge higher interest rates for minorities. In addition, facial recognition technologies have the highest error rates among Black women. If left unchecked, biased algorithms will further exacerbate discrimination, inequality, and racism.

A copy of the legislation can be found HERE.

Cosponsors in the Senate include Senators Cory Booker (D-N.J.), Ben Ray Lujan (D-N.M.), Jeff Merkley (D-Ore.), Elizabeth Warren (D-Mass.), Peter Welch (D-Vt.), and Ron Wyden (D-Ore.). 

The Eliminating BIAS Act is endorsed by Lawyers’ Committee for Civil Rights Under Law, Leadership Conference on Civil and Human Rights, Center for Democracy and Technology, National Urban League, Electronic Privacy Information Center (EPIC), Free Press Action, Public Knowledge, Accountable Tech, Demand Progress, Fight for the Future, Common Sense Media, Center for Digital Democracy, Common Cause, Open Technology Institute, Upturn, National Hispanic Media Coalition, Asian Americans Advancing Justice – AAJ, UnidosUS, The Trevor Project, National Action Network, Fairplay, National Urban League, National Council of Negro Women, and Access Now. 

“Algorithmic systems profoundly influence opportunities and outcomes throughout society. Black people and other communities of color disproportionately suffer negative impacts of unjust and opaque AI systems. The Eliminating BIAS Act is an important step towards fostering transparency, oversight, and justice for emerging technologies. By requiring each agency to report on harms from algorithmic systems, this bill would incentivize agencies to identify, prevent, and mitigate threats to civil rights from AI,” said David Brody, Managing Attorney of the Digital Justice Initiative at the Lawyers’ Committee for Civil Rights Under Law.

“The civil rights community has long advocated for the need to both establish new offices of civil rights across the federal government and provide increased support for existing offices. Senator Markey’s Eliminating BIAS Act does exactly that by ensuring that every federal agency that uses, funds, or oversees artificial intelligence has an office of civil rights focused on combating bias, discrimination, and other harms. It’s imperative that the federal government keep pace with emerging civil rights threats stemming from technological innovation powered by AI. This legislation takes an important first step in developing that infrastructure,” said Koustubh “K.J.” Bagchi, Vice President at the Center for Civil Rights and Technology at The Leadership Conference on Civil and Human Rights.

“There is currently little oversight as to whether artificial intelligence systems are producing results that are accurate, fair, and unbiased, and federal agencies are in a unique position to evaluate and publish information about how AI is used in the sectors they regulate. Senator Markey’s Eliminating BIAS Act puts in place a commonsense requirement that agencies have an office of civil rights to evaluate AI systems and help minimize harms stemming from their use. EPIC is proud to support this legislation,” said Caitriona Fitzgerald, Deputy Director of the Electronic Privacy Information Center.

“The use of artificial intelligence has proliferated in our digital landscape. It is critical that the federal government keep abreast of this prevalent issue to effectively evaluate and address the emerging harms of algorithmic bias that target historically marginalized and underserved communities. NHMC commends Senator Markey for taking the initiative to ensure our civil rights are protected in an ever-changing digital age,” said Brenda Victoria Castillo, President and CEO of the National Hispanic Media Coalition.

“It is critical that civil rights considerations be at the heart of U.S. technology policy, and the Eliminating BIAS Act represents a meaningful step towards that goal. The proposal advocates for practical, common-sense measures, like establishing a civil rights office with experts and implementing comprehensive reporting mechanisms to mitigate harm. This aligns seamlessly with civil society's demands, ensures federal agencies embrace transparency and protects against deploying AI systems that can violate peoples’ rights,” said Willmary Escoto, U.S. Policy Counsel at Access Now.

“New America’s Open Technology Institute is pleased to endorse the Eliminating BIAS Act, which requires federal agencies to establish civil rights offices focused on bias, discrimination, and other harms that may arise from the use of algorithms. This proposed legislation complements Executive Order 14110’s focus on centering rights and fairness in the federal government’s development, use, and oversight of artificial intelligence tools. Empowering civil rights offices within federal agencies is critical to ensuring that artificial intelligence is used responsibly and in the public interest,” said Prem Trivedi, Policy Director at New America's Open Technology Institute.

“Rapid advances in AI, the frenzied deployment of new systems, and the surrounding hype cycle have generated a swell of excitement about AI’s potential to transform society for the better. But concrete harms from these systems are already being felt, and are advancing as rapidly as AI itself – exacerbating bias and discrimination, eroding privacy and autonomy, and more. It is critical to bring experts and technologists into the conversation to create standards for how algorithms are deployed under the jurisdiction of all federal agencies – especially as specific standards for evaluation and documentation will necessarily differ depending on the type and maturity of the AI. We commend Senator Markey’s leadership on this issue and urge Congress to take immediate action,” said Nicole Gill, Executive Director at Accountable Tech.

“Artificial intelligence should not be used as an excuse to skirt the law, but without accountability that is what will happen. It is critical to have a functioning society that the federal government is equipped to do their job, which in the 21st century means overseeing artificial intelligence and ensuring that Americans’ civil rights are protected. We commend Senator Markey’s Eliminating BIAS Act for his leadership in addressing this critical gap in the federal government at this important time,” said Ishan Mehta, Director for Media and Democracy at Common Cause. 

“Free Press Action welcomes Senator Markey’s important and ongoing focus on the risks that machine learning tools pose to civil rights in the United States. There’s mounting evidence that these tools can cause disproportionate harm to Black and brown communities, including in critical contexts like immigration, policing, and healthcare. Government agencies urgently need increased capacity to anticipate, measure and mitigate civil rights harms stemming from the adoption of artificial intelligence tools to perform key societal functions. This legislation takes a vital step toward safeguarding vulnerable communities from harms stemming from misuse of these technologies,” said Jenna Ruddock, Policy Counsel at Free Press Action.

“The idea is simply commonsense: Federal agencies that oversee people’s rights and economic opportunities should have civil rights offices — and it’s imperative that these offices have the staff capacity to examine the algorithms that affect people’s lives,” said Harlan Yu, Executive Director at Upturn.

“While AI and emerging technologies have the potential to revolutionize the provision of critical services at the federal level, they also hold immense potential to impact marginalized communities — including LGBTQ+ populations — both positively and negatively,” said Casey Pick, Director of Law and Policy at The Trevor Project. “Without the right safeguards and measures in place, algorithmic bias can harm underrepresented groups of people. That’s why it’s so important to put in place protections from discrimination and bias early in the adoption of new technologies, particularly given the significant impact that government programs can have in addressing disparities in areas like healthcare access for LGBTQ+ communities.”

Senator Markey has called on the federal government to hold Big Tech accountable, investigate AI, and stop algorithmic injustice. Last month, Senator Markey sent a letter to Clearview AI demanding answers from the company on its facial recognition technology.?In October, he and Congresswoman Pramila Jayapal (WA-07) applauded the Biden administration for heeding their call to incorporate the White House Blueprint for an AI Bill of Rights into its AI Executive Order. In July, Senator Markey and Congresswoman Doris Matsui (CA-07) reintroduced their Algorithmic Justice and Online Platform Transparency Act to ban discriminatory algorithms and improve transparency on social media platforms. In March, Senators Markey and Jeff Merkley (D-Ore.), along with Congresswoman Jayapal,?led?their colleagues in reintroducing the?Facial Recognition and Biometric Technology Moratorium Act?to prevent the government from using facial recognition and other biometric technologies.?