Race and technology in America

National

The technology industry is famously determined to change the world — but its efforts to diversify its workforce and remove bias from its products haven't changed nearly enough. 

The technology industry is famously determined to change the world — but its efforts on race haven't changed nearly enough.

Illustration collage of an old computer monitor surrounded by graphic shapes

Deep Dive: Hard Truths - Technology

The technology industry is famously determined to change the world — but its efforts to diversify its workforce and remove bias from its products haven&#x27;t changed nearly enough.

From the Valley to D.C., Big Tech players like Facebook, Google and Amazon are under more scrutiny than ever as new technology develops and privacy and antitrust concerns grow in lockstep with companies’ ambitions.

Technology

Axios AI+

The tech industry likes to cast its failure to include enough people of color and other underrepresented groups as a "pipeline problem" — one that would vanish if only more such people studied tech skills and entered the field. 

But there's another reason U.S. tech companies struggle to diversify: work environments that critics say are rife with harassment and discrimination even as companies paint themselves as champions for diversity.

Why it matters: The hard truth is that filling up the pipeline won't help, as long as employees keep quitting their jobs, being forced out or leaving the industry. 

By the numbers: Companies do a better job documenting overall representation by race than they do at reporting attrition rates, which aren't always included in publicly released diversity reports.

Google is the exception. Its report shows attrition among Black employees was 12% higher than the overall rate in 2020 and 21% higher so far in 2021. 

Those numbers aren't conclusive — the company also lost white employees at a higher than average rate, while retaining Asian employees in higher numbers. 

One extensive study of tech attrition, conducted by the Kapor Center in 2017, found that "employees within tech companies experienced significantly more unfairness than employees in non-tech companies" and that "underrepresented men of color were most likely to leave due to unfairness."

Departing employees, experts note, will often simply leave a job without talking about experiences of discrimination, believing they won't be heard or concerned that they'll jeopardize their next job hunt. 

Few have the emotional and financial wherewithal to take on a big tech company. "Many people can’t afford to take the risk of having no health insurance, having no income," said Ifeoma Ozoma, a former Pinterest employee and whistleblower. "It is a very personal decision to speak up. It’s also not easy for those who do."

Yes, but: Current and former employees continue to sound alarms on social media.

Google associate product manager Angel Onuoha recently tweeted about the experience of having security called on him as a suspicious person while simply working and biking on campus. The tweet went viral, and others shared similar experiences at Google and elsewhere.

A growing number of tech employees who raise issues of discrimination say that part or all of the company's reaction has been to suggest counseling or medical leave, rather than addressing the alleged discriminatory conduct.

Rashad Robinson, president of the activist group Color of Change, said that Black women and women of color have borne the brunt of harassment in tech and have been among the most willing to speak out, risking their next job in the field. 

Companies often react to such issues as if this is the first time they are hearing about them, even as they have heard such stories over and over, Robinson said. "It doesn't even actually pass the smell test."

Pinterest paid $22.5 million to settle a gender pay discrimination suit at the end of last year. However, that settlement was made with the white COO who pursued the issue and not the two Black women — Ozoma and Aerica Shimizu Banks — who earlier had raised issues of racial and gender discrimination. 

Girls Who Code CEO Tarika Barrett said that when she talks with girls considering working for a tech company, she tells them — especially young women of color — that they are likely to find that they are the only one like themselves in the room.

She also noted a recent survey of Girls Who Code participants who were interning at tech companies that found half of the respondents had either personally experienced harassment or knew someone who had.

The big picture: In tech, the failure to represent groups inside companies also influences the products they produce, baking in harmful biases.

The bottom line: Leaders in the field say many of the social problems caused by tech products could have been identified early and even corrected if their creators came from more-diverse backgrounds.

Robinson cites Zoom's failure to anticipate the need to protect meetings from malicious disruptions: "All you had to do was talk to some Black folks."

Barrett says the high stakes are why it's so critical that people of color are represented at tech companies. "There is no opting out of tech."

Go deeper: Tech's race problem is all about power

Fixing the pipeline won't solve tech's race problem

Increasing the supply of new workers doesn't help if underrepresented employees keep leaving.

Illustration of a recursive tunnel of hands holding smart phones.

The tech industry likes to cast its failure to include enough people of color and other underrepresented groups as a &quot;pipeline problem&quot; — one that would vanish if only more such people studied tech skills and entered the field. But there&#x27;s another reason U.S. tech companies struggle to diversify: work environments that critics say are rife with harassment and discrimination even as companies paint themselves as champions for diversity.

Data: Google, Apple, Facebook, Microsoft, Amazon; Chart: Jared Whalen/Axios

In the mid-2010s tech's biggest companies started publicly reporting diversity statistics, and the tale these annual numbers tell since then is one of progress — but very slow and very slight.

U.S. Census numbers for 2020 show the U.S. Black population at 12.4% and the Hispanic population at 18.7%. These numbers don't include the growing tally of multi-racial people. 

By the numbers: Tech's trickle of change

Tech giants' self-reported racial representation numbers have improved, but only very slowly.

Emerging Tech

In the mid-2010s tech&#x27;s biggest companies started publicly reporting diversity statistics, and the tale these annual numbers tell since then is one of progress — but very slow and very slight.U.S. Census <a class="gtmContentClick" data-vars-link-text="numbers for 2020" data-vars-click-url="https://www.census.gov/library/stories/2021/08/improved-race-ethnicity-measures-reveal-united-states-population-much-more-multiracial.html" data-vars-content-id="adf91373-aa1d-44e6-ac34-887cc8064b64" data-vars-headline="By the numbers: Tech&#x27;s trickle of change" data-vars-event-category="story" data-vars-sub-category="story" data-vars-item="in_content_link" href="https://www.census.gov/library/stories/2021/08/improved-race-ethnicity-measures-reveal-united-states-population-much-more-multiracial.html" target="_blank">numbers for 2020</a> show the U.S. Black population at 12.4% and the Hispanic population at 18.7%. These numbers don&#x27;t include the growing tally of multi-racial people.

Gaming

Axios Gaming

Game developer Chandana "Eka" Ekanayake waited till mid-career to found a startup because he didn't see a path to do it. "I joke about this on Twitter: I wish I had 'white man confidence,'" he says now. "I would have started a company at 25."

Historically, that confidence hasn't been encouraged in the gaming sector. Ekanayake is now one of several entrepreneurs aiming to speed up diversification in a field where progress has been slow.

By the numbers: A quarter of the 200+ million video games players in the United States are people of color, according to the Entertainment Software Association.  But the executive ranks of the industry, as well as rank-and-file game development and the power players in games media, are less diverse.

One measure: The group Black in Gaming says that only 2% of developers are Black.

Another: A recent leak of Twitch data showed that the list of best-paid streamers is dominated by white men. 

After the killing of George Floyd last year, media CEO Regi Cash heard from the young workers at his digital entertainment company that they wanted to make a difference. Cash agreed and added a key goal: making that difference last.

“We needed to do something in a more sustainable fashion,” Cash told Axios of his team's project, “if we really wanted to have an impact.”  

Needed change: Cash and his team at 3BlackDot are now creating a series of video shows called “Gaming While Black,” which are designed to attract both Black gamers and Black talent in front of and behind the camera.

The project came from a desire to uplift Black voices after Floyd’s killing, but it's been shaped by Cash’s recognition of where his own company, which prided itself on diversity, had fallen short. 

Cash realized that 3BlackDot was paying out $5 million a month to content creators it had partnered with, but was only distributing 3% of that to creators of color. 

Without a more intentional effort, they hadn’t found a diverse talent pool.

To solve that, Cash and his team took a more systemic approach. 

He explored the reasons for a lack of visible Black content creators. For one, lack of representation was a barrier. "I firmly believe that you have to see yourself in it to believe that it's possible for you," Cash said.

He also identified a financial catch-22. The industry didn't invest in Black talent, so Black content creators couldn't build an audience and attract sponsors — unless those sponsors put money in to make the audience grow.

Cash's project would need to not only find an audience, but find talent, find distribution and prop up an ecosystem in which Black creators and others could succeed.

For Ekanayake, the challenge of making gaming a more sustainably diverse space has been a long, personal journey. 

Emigrating from Sri Lanka at 8 and starting in the industry at 19, he worked for major studios for nearly two decades before co-founding his own, Outerloop, in 2017.

For most of his career he didn’t emphasize his cultural background. “I was like, I got to try to fit in,” he told Axios. “I was always worried about losing my job.”

Age and parenthood made him rethink his background and his interests, and he began to dream of starting a studio that would elevate themes from other cultures.

Dreams come true more easily with financing, and that’s where a fluke experience with virtual reality helped Ekanayake start Outerloop.

About eight years ago, VR became the rage, and investors were leaping at a chance to fund VR games. 

Ekanayake, working for a studio called Uber Entertainment at the time, got the chance to pitch and lead some VR gaming projects. That gave him the opportunity to learn how to pitch, secure funding and run a team.

His epiphany: It isn’t that hard if you have the confidence, and if you’re given the chance to take that chance. 

Video game screenshot of a person holding a baby falcon on their arm

In 2019, Outerloop released its debut game, "Falcon Age," which mixes a simulation of falconry with a fight against colonizers.

It was enough of a success to expand Outerloop's diverse team and spur a new, mystery project.

"My success level is, 'Oh, we didn’t lose money,'" he said, "We have enough to do another one."

The big picture: For Cash, Ekanayake and others, they’ve found that building projects that support diversity also means doing work on the foundation below it. 

Ekanayake acknowledges that's a "harder road," but said he's inspired because he's working in a creative industry and trying to pay his lessons forward to other aspiring developers of color.

That, he reasons, can help the next generation and make this progress last.

Video gaming's "harder road" to racial diversity

A handful of entrepreneurs are setting out to speed up diversification in a field where progress has been slow.

Photo illustration of a screen displaying a Pong game half inverted. 

Photo illustration: Aïda Amer/Axios. Photo: Barry James Gilmour/Fairfax Media via Getty Images

Game developer Chandana &quot;Eka&quot; Ekanayake waited till mid-career to found a startup because he didn&#x27;t see a path to do it. &quot;I joke about this on Twitter: I wish I had &#x27;white man confidence,&#x27;&quot; he says now. &quot;I would have started a company at 25.&quot;Historically, that confidence hasn&#x27;t been encouraged in the gaming sector. Ekanayake is now one of several entrepreneurs aiming to speed up diversification in a field where progress has been slow.

Asians make up the majority of Silicon Valley's tech workforce at roughly 57%, according to MarketWatch. Yet they're vastly underrepresented at the leadership level: 27% at Apple, 40% at Google and 25% at Facebook.

The big picture: Many tech leaders like to think the field is "post-race," often pointing to the handful of Asians, mostly East and South Asian men, who occupy prominent executive roles. The reality is far more complicated.

Why it matters: The industry’s rhetoric around Asians has long obscured disparities that affect Black and Latino workers. It also overlooks Asians’ precarious place in the system — and tech's long history of anti-Asian bias.

State of play: White-collar Asian American professionals are the least likely racial group to be promoted into management in Silicon Valley, according to a 2017 analysis of national workforce data that was published in the Harvard Business Review.

"All of that comes down to the stereotypes around what leaders look like and … a lot of assumptions about Asian characteristics," said Ellen Pao, who made national news in 2012 when she filed a gender discrimination suit against the Silicon Valley venture capital firm Kleiner Perkins Caufield & Byers.

Yeong Cheng, a software management executive and the founder of Denver Asian Collective, said they've had to fight to get to where they are in their career.

After unofficially managing a full-time team at a computer software company for six months, Cheng said they were denied a promotion to manager even though the six white men who started with Cheng all got bumped to management. 

Even in the last few years serving in executive roles, they say they're often infantilized. "I have been second-guessed as a default, until I've proven beyond a very high bar that not only do I know what I'm talking about, but I'm the expert in the room," they told Axios.

Every level of work is saturated with these biases, whether based on Asians’ accents or names, cultural differences, a perception of their foreignness or racial stereotypes that pigeonhole Asians as submissive or robotic.

Asian women face additional barriers and burdens, as reflected in the flurry of gender and race discrimination lawsuits in the tech industry that were triggered by Pao's. 

Case in point: A former employer once told Pao he only wanted to have Asian women work for him because they "work so hard and don't complain," she said.

The backstory: Asian Americans have made up a good chunk of the tech industry since the 1960s, when the U.S. repealed laws largely excluding Asians from entering the country. 

At the time, the government prioritized entry for Asians with educational backgrounds or occupational specialties — primarily from East and South Asia. 

Their status in America was largely conditional, however — dependent on the work they produced and their willingness to assimilate. 

They were also easily exploited by white employers who saw them as "good work horses" but not leaders, as engineer David K. Lam told the New York Times in a 1992 article. 

The big picture: According to scholar Claire Kim's racial triangulation theory, America's race hierarchy places Asians between superior and inferior but solidly labels them foreign.

That foreignness is a constant shadow, yet for decades, Asian Americans heard the same refrain: "You should be grateful."

The industry’s discussions around diversity have typically excluded its prominent Asian worker population, a move that implies they face no adversity since they seemingly have a foot in the door, said Association for Asian American Studies president-elect Pawan Dhingra.

Yet 44% of Asians in STEM jobs report experiencing discrimination at work, according to a 2018 report from Pew Research Center.  

What they’re saying: "There's a huge problem around pitting groups against each other instead of trying to fix the system," Pao told Axios.

When Cheng has tried to raise concerns about anti-Asian racism, their colleagues have told them they're "distracting from [Black Lives Matter]." 

The bottom line: Pao, who launched the nonprofit Project Include to help companies improve on equity, said that amid a rise in anti-Asian hate and U.S.-China tensions, "we're at this unique point where we have an opportunity to push back on racism or allow it to continue to fester."

"We're definitely having people understand what's going on better," she added. "The hard part is getting people to actually act on it." 

Editor's note: We've updated the identification of Yeong Cheng in this story. 

Tech's troubled history with Asian workers

Asians make up the majority of Silicon Valley's tech workforce, but are underrepresented in leadership.

Illustration of a laptop with a woman on the screen surrounded by circles and shapes

Big Tech

Asians make up the majority of Silicon Valley&#x27;s tech workforce at roughly 57%, <a class="gtmContentClick" data-vars-link-text="according to MarketWatch" data-vars-click-url="https://www.marketwatch.com/story/asian-american-tech-leaders-rally-against-wave-of-hatred-we-dont-deserve-to-live-in-fear-in-our-own-country-11617740339" data-vars-content-id="73ea8af9-d522-4144-b376-a92347016a95" data-vars-headline="Tech&#x27;s troubled history with Asian workers" data-vars-event-category="story" data-vars-sub-category="story" data-vars-item="in_content_link" href="https://www.marketwatch.com/story/asian-american-tech-leaders-rally-against-wave-of-hatred-we-dont-deserve-to-live-in-fear-in-our-own-country-11617740339" target="_blank">according to MarketWatch</a>. Yet they&#x27;re vastly underrepresented at the leadership level: 27% at Apple, 40% at Google and 25% at Facebook.The big picture: Many tech leaders like to think the field is &quot;post-race,&quot; often pointing to the handful of Asians, mostly East and South Asian men, who occupy prominent executive roles. The reality is far more complicated.

Macro

Axios Macro

Tech's leading companies — like the rest of corporate America — responded to the murder of George Floyd by police in 2020 in part by pledging to give big to racial justice groups.

Our analysis of these companies found that, by and large, Big Tech paid up.

Why it matters: Amazon, Apple, Facebook, Google and Microsoft have deep enough pockets to make substantial donations and other aid adding up to hundreds of millions of dollars.

Yes, but: Critics of philanthropy say giving can be a way to paper over failures to make real change happen inside boardrooms and workplaces.

Between the lines: Many of the companies have also promised to improve conditions internally for Black employees and prospective hires. The high-dollar donations provide one external gauge to track commitments to change.

Of note: Donations are gifts, but the portion of these commitments that takes the form of investment can have a financial upside for the companies.

What it promised: Shortly after Floyd was killed, Google.org pledged $12 million in donations to racial justice groups, which grew to $320 million in commitments from Google and YouTube in the last year.

How it works: These pledges took the form of direct donations, ad grants and technical support. Google also invested $60 million in Black-led startups and venture capital firms, with plans to invest another $40 million by the end of the year.  

What they did: The Center for Policing Equity was one of the first groups Google said would receive a $1 million donation.

Through a partnership with Google.org, Google’s philanthropy arm, CPE launched Justice Navigator in September, which provides data on use of force incidents and vehicle and pedestrian stops, with recommendations on how to use the assessments to advocate for change.

What it promised: Microsoft says it has committed $246 million toward racial justice efforts since Floyd’s murder, including $76 million in grants and $170 million in loans, investments or deposits. 

Separately, the company also said last year it plans to double the number of Black-owned suppliers in its network in three years and spend an incremental $500 million with the new and existing suppliers.

How it works: Microsoft said it would enhance the weighting of diversity characteristics during the supplier evaluation and selection process, as well as push its top 100 suppliers to make progress on Black representation in their own workforces. 

What they did: In an update earlier this year, Microsoft said it had “realized growth” in its Black-owned supplier base and implemented a confidential, voluntary workforce representation survey.

The company declined to disclose specific spends or supplier amounts so far.

What it promised: Apple announced a long-term $100 million commitment through a new Racial Equity and Justice Initiative and expanded it to include an additional $30 million this year to help communities of color, with a focus on education, economic equality and criminal justice reform.

How it works: So far, REJI has provided funding to historically Black colleges and universities and Hispanic-serving institutions, created an app developer academy in Detroit, and invested $50 million in venture capital firms focused on minority-owned businesses.

What they did: The Apple Developer Academy in Detroit opened to its first 100 students this month, offering a free, 10-month program to teach coding, design and entrepreneurship — within the Apple ecosystem. Apple expects to expand the program to 1,000 students.

Our thought bubble: That's a lot of potential new developers for Apple’s App Store.

What it promised: Amazon said it would donate $27 million (including employee contributions and company matches) to 12 groups “focused on combating systemic racism,” like the Thurgood Marshall College Fund and the National Urban League.

The company — one of the world’s largest employers — also said it set goals to double the number of top Black executives in the wake of Floyd’s murder.  

How it works: The National Urban League said it received over $1 million from Amazon. A spokesperson from the group said the donations were unrestricted, meaning they can go toward any of its programs and operations. Amazon pledged another unrestricted $750,000 to the organization for 2022.

What they did: Amazon met its hiring goal, according to its workforce data. The improvement, though, is off a low base: In 2019, just 1.9% of senior leaders were Black. Now that's 3.8%.

What it promised: Facebook said it would give $10 million to more than 30 racial justice groups, including All Star Code and the Shriver Center on Poverty Law. It also pledged to invest $100 million in Black-owned small businesses.

It said it would distribute 100,000 scholarships to Black students in its Facebook Blueprint program, which credentials digital advertisers.

What they did: Of the $100 million commitment to Black-owned businesses, the company said in an update this year that it has so far distributed $40 million of that in the form of grants and ad credits.

Big Tech's donation dollars

Companies rushed to donate money in the aftermath of George Floyd's killing. Here's how they've delivered so far.

Photo illustration of a statue of George Floyd with money falling around it. 

Photo illustration: Aïda Amer/Axios. Photo: Alexi Rosenfeld/Getty Images

Tech&#x27;s leading companies — like the rest of corporate America — responded to the murder of George Floyd by police in 2020 in part by pledging to give big to racial justice groups.<ul><li>Our analysis of these companies found that, by and large, Big Tech paid up.</li></ul>

Latino

Axios Latino

From Palantir to Ring, to body cameras and GPS databases, technology has transformed policing — and incorporated the biases of its creators.

Why it matters: These advances can fight but also foster racial discrimination.

Details: The death of George Floyd at the hands of Minneapolis police was caught on a teen's cell phone video. Images of a white officer kneeling on Floyd's neck sparked worldwide protests and a national reckoning.

The video went viral on various social media platforms as it was viewed and shared millions of times.

Darnella Frazier was later awarded a special citation by the Pulitzer Prize Board.

Yes, but: Law enforcement agencies also use technology that critics say reinforces racial bias.

Facial recognition: Law enforcement agencies have long used forms of facial recognition technology, but in recent years better technology has been adopted by departments in New York, Los Angeles, Chicago and other cities.

Forms of biometric security can confirm an individual’s identity using faces in photos and videos or in real time. A 2016 study by the Center on Privacy & Technology at Georgetown Law found that 1 in 2 American adults is in a law enforcement face recognition network.

But activists say the technology can misidentify people with dark skin and violate privacy rights. A coalition of nearly 70 civil rights and advocacy groups in June urged Congress to ban the technology in some places.

Surveillance apps: Home security products, like Amazon's Ring video doorbells and Google's Nest, allow residents to record images and video footage at their doorsteps but also from nearby sidewalks and streets.

Ring’s affiliated crime reporting app, Neighbors, allows people to upload images that sometimes misidentify people of color walking dogs.

Users have also posted racist language to describe Black and Latino residents.  While the Neighbors app bans harassment and bullying, it appears impossible to stop these problems effectively through existing monitoring.

Ring has partnerships with around 1,800 U.S. police departments that can request camera footage from Ring doorbells without a warrant. The company declined to say how many users had footage obtained by police.

The American Civil Liberties Union said such a system amounts to spying on residents and called growing partnerships between such apps and law enforcement troubling.

Real-time crime centers: A growing number of police departments are launching surveillance hubs so officers can use an expanding range of GPS, monitoring and database technologies. 

The centers work as central nodes and control rooms for automated license plate readers, gunshot detection, real-time social media monitoring, predictive policing algorithms and a network of video cameras.

The Atlas of Surveillance project, the Electronic Frontier Foundation and the Reynolds School of Journalism at the University of Nevada, Reno have identified more than 80 such centers in 29 states.

Critics say the centers allow police to needlessly monitor peaceful protests, bicycle races or other large gatherings by people of color.

“They basically can record all of our associations and where we go, who we see, where we slept, over time because they record all the location data,” Tessa D’Arcangelew, organizing program manager at the ACLU of Northern California, told KTXL-TV in Sacramento, California.

Policing the police: Some mobile phone apps seek to stop rather than simply document deadly police encounters with people of color — including notifying family and lawyers about potential violations in real time.

The Cop Watch Video Recorder app opens with Siri on iPhones, automatically beginning filming and sending footage to the cloud. The app films in real time in case the police officer seizes or breaks the phone.

The Legal Equalizer app captures police encounters after the user is pulled over, automatically notifying loved ones and providing basic legal information on the spot.

The Mobile Justice app, available in all 50 states, records and submits police incidents directly to local chapters of the ACLU. The app also lets users send videos via text messaging to family and private attorneys.

Technology, policing and racial bias

These advances can fight but also foster racial discrimination.

Illustration of a person's face under scrutiny from facial recognition technology

Latin America and Latinos

From Palantir to Ring, to body cameras and GPS databases, technology has transformed policing — and incorporated the biases of its creators.Why it matters: These advances can fight but also foster racial discrimination.

Future

Axios Future

AI algorithms employed in everything from hiring to lending to criminal justice have a persistent and often invisible problem with bias.

The big picture: One solution could be audits that aim to determine whether an algorithm is working as intended, whether it's disproportionately affecting different groups of people and, if there are problems, how they can be fixed.

But the field of algorithmic auditing is still in its infancy, and until the AI field is governed by meaningful regulations, it will be challenging to carry out audits worthy of the name.

How it works: Algorithmic audits — usually conducted by outside companies — involve examining an algorithm's code and the data used to train it, and assessing its potential impact on populations through interviews with stakeholders and those who might be affected by it.

But unlike financial audits — which have clearly defined objectives codified in established law — the field of algorithmic auditing "is a bit disjointed," says Liz O'Sullivan, CEO of the AI auditing company Parity.

"There are a lot of different companies that are trying to do different kinds of approaches to validating whether a model is performance compliant or discriminatory," she adds.

Between the lines: Financial audits exist in part to open up the black box of a company's internal operations to outside investors, and ensure that a company remains in compliance with financial laws and regulations.

In the case of algorithmic audits, however, the actual workings of AI can be a black box to the company itself because unless explainability is built into the foundation of an algorithmic model, it can be easy to get lost.

That can make it difficult for auditors to determine how and where an algorithm might be going wrong, let alone "translate that mathematical representation into words that a general person can understand," says Henry Jia, data science lead at the software company Excella.

Untangling that process often involves going back to the original data. "A majority of the biases that become part of the model is inherited from problems in the data," adds Jia.

Details: Algorithmic audits can help companies screen their AI products for flaws that may not be apparent at first glance.

Utah-based HireVue, which uses AI to help screen millions of job applicants as part of its recruitment services, contracted with outside auditors to evaluate its hiring process for early-career applicants.

The audit turned up issues with how the model can handle different accents in processing video interviews, notes Lindsey Zuloaga, HireVue's chief data scientist.

Yes, but: HireVue was criticized by some outside observers who accused the company of using the results of the audit as marketing material rather than an opportunity for fundamental change. (HireVue says it was transparent with the audit by making the full report public.)

Part of the problem, notes O'Sullivan, is that outside of specific areas like lending and hiring that have well-established anti-discriminatory regulations, many of the fields now touched by algorithms are essentially "like the wild, wild West."

Audits can help, but more fundamental change will require new laws around AI and the updating of regulations put in the books decades before anyone was working with algorithms.

"We have this incredible new vocabulary of fairness, new ways of thinking about it, that are simply not in the law."

What to watch: Whether the push to regulate Big Tech leads to laws specifically designed to hold algorithms to account, like the 2019 Algorithmic Accountability Act, which would have required companies to audit their algorithms and address any revealed bias issues before they're put into use.

Audits attempt to clean up AI bias

But the field needs meaningful laws and regulations before it can have a meaningful effect

Illustration collage of a magnifying glass over binary code

Artificial intelligence

Artificial Intelligence

AI algorithms employed in everything from hiring to lending to criminal justice have a persistent and often invisible problem with bias.The big picture: One solution could be audits that aim to determine whether an algorithm is working as intended, whether it&#x27;s disproportionately affecting different groups of people and, if there are problems, how they can be fixed.

Race and media in America

This 13th Axios AM Deep Dive in our Hard Truths series explores race and the media industry.

Illustration of a video camera surrounded by a film frame. 

This 13th Axios AM Deep Dive in our Hard Truths series explores race and the media industry.

Share this Deep Dive

Race and technology in America

Next Deep Dive

Race and media in America