President Joe Biden‘s new artificial intelligence executive order sets standards for diversity and equity, drawing praise from liberals groups who said it will help protect vulnerable groups but condemnation from conservatives, who warned it will advance controversial left-wing objectives.
Biden’s Oct. 30 order calls on the federal government to begin to gather resources, develop understanding, and work toward bringing sufficient experts in to assist in using and regulating AI. Most of the order focused on how to manage the technology responsibly, but a thread throughout the order is investigating how to mitigate biases inadvertently programmed into AI.
GOOGLE GOES TO COURT WITH FORTNITE DEVELOPER OVER APP PRICING DISPUTES
“My Administration cannot — and will not — tolerate the use of AI to disadvantage those already too often denied equal opportunity and justice,” Biden said in the order.
Setting Guidelines
The order asks that the attorney general organize a meeting with various agency leads to discuss discrimination in AI.
This includes collaborating with the agencies to compile a report addressing how AI could factor into the legal process, such as through sentencing recommendations, parole, the processing of prisoners, and risk assessments. The meeting is also meant to analyze how the bias that scholars allege is in the software algorithm may affect police surveillance, crime forecasting, and forensic analysis. An ensuing report would then be used to formulate best practices and provided to other agencies for their adoptions.
The attorney general will also collaborate with the secretary of homeland security to develop recommendations for state, local, and tribal law enforcement agencies so that they can hire people skilled with the technology and able to use it appropriately.
The health and human services and agriculture secretaries will have 180 days to issue guidance for their representative agencies, including public benefits and services.
Combating discrimination
Federal agencies will also have to use their civil rights offices to prevent unlawful discrimination based on the decisions made by AI systems. This includes providing more precise guidelines and additional communication between offices on how to use AI appropriately.
The secretary of labor was also ordered to publish guidance for hiring that restricts discrimination based on AI or other technology-based hiring systems. It will also address the racial bias in the housing and consumer financial markets, including by addressing and restricting any use of screening software that one might consider a violation of the Fair Housing Act or other related federal laws.
Supporters and critics
Civil rights groups praised the guidelines presented by Biden.
“Artificial intelligence has become integrated into our daily lives in significant, but often subtle ways, exacerbating and magnifying discriminatory harms in housing, education, employment, and more,” Cody Venzke, senior policy counsel in the American Civil Liberties Union’s National Political Advocacy Department, said in a statement. “Today’s executive order invokes existing civil rights authority to address discriminatory AI throughout the government.”
Maya Wiley, president and CEO of the Leadership Conference on Civil and Human Rights said in a statement that “we still need Congress to consider legislation that will regulate AI and ensure that innovation makes us more fair, just, and prosperous, rather than surveilled, silenced, and stereotyped.”
Conservatives, though, who have become increasingly skeptical of initiatives under the banner of “equity,” criticized the order.
“President Biden has signed an executive order that will require AI companies to ‘address algorithmic discrimination’ and ‘ensure that AI advances equity,'” Christopher Rufo, senior fellow at the Manhattan Institute, posted. “They want to embed the principles of [critical race theory] and [diversity, equity, and inclusion] into every aspect of AI.”
“Just imagine the possibilities: race-based government benefits, tax audits, and criminal prosecutions, all advanced through the government’s preferred AI platforms, but no human actually makes the decisions,” Gene Hamilton, general counsel at America First Legal, told the Washington Examiner. The Biden administration will use AI as “a blunt weapon to advance their extreme goals,” Hamilton said. AFL is a right-leaning organization founded by former Trump official Stephen Miller.
Biden previously signed an executive order in February that Rufo claimed would create “woke” AI. The order stated that agencies develop AI systems “in a manner that advances equity.”
CLICK HERE TO READ MORE FROM THE WASHINGTON EXAMINER
Some scholars have argued that AI models suffer from human biases that are unintentionally put into the models by their programmers and designers. These biases may lead to unethical and discriminatory decisions based on race or gender.
Biden’s order is an expansive effort that will begin to provide guidelines on how to implement AI safety and how to ensure user privacy in the process. One early criticism of the order is that Biden may not have the necessary resources for proper enforcement.