ethics
People

'There's a chilling effect': Google's firing of leading AI ethicist spurs industry outrage

Timnit Gebru's firing could damage Google's reputation and ethical AI research within tech companies, industry leaders told Protocol.

'There's a chilling effect': Google's firing of leading AI ethicist spurs industry outrage

Timnit Gebru said that she was forced out of Google because of an email she sent to members of the Google Brain Women and Allies listserv.

Photo: Kimberly White/Getty Images

After Google fired one of the industry's most respected and well-loved AI ethics researchers on Wednesday, Google employees and tech industry leaders alike voiced their fear that her firing will have a "chilling effect" on ethics research within tech companies and at Google specifically.

Timnit Gebru, the now-former technical co-lead for Google's AI ethics team, said that she was forced out of Google because of an email she sent to members of the Google Brain Women and Allies Listserv that detailed her frustration with the company's diversity pledges and the exhausting experience of being a Black woman at Google, as well as conflict over an ethics research paper that Google wanted retracted. Over the last week, Gebru had been fighting to have the research paper — which discusses the ethics issues of large language models — published with her and other Google employees' names.

After Gebru said that she would plan to resign if Google didn't commit to further discussion about the company's demands over the research paper, Google immediately rejected her conditions and terminated her employment without discussion, according to Gebru's statement. In the email explaining her termination shared by Gebru, Google Research Vice President Megan Kacholia wrote that Gebru's email to the Listserv was "inconsistent with the expectations of a Google manager." Google declined to comment.

Gebru is best known for her research on discrimination within facial recognition models, including a groundbreaking study that illustrated gender and skin-type bias in the best commercial AI facial recognition systems at the time. "She's literally the best of the best. She's the best that we've got. Not only does Timnit encapsulate our hopes and dreams, and is the embodiment of the best of us, but she is strongly supported," said Mutale Nkonde, the CEO of AI for the People and a fellow at Harvard's Berkman Klein Center.

Gebru is also well-liked for supporting activism within Google and defending employees who've lost their jobs because of their protests. Shortly before she tweeted that she had been fired, the National Labor Relations Board filed a complaint that said Google had violated labor laws by spying on and then firing workers who were organizing employee protests. "If we have heroes in the AI ethics community, she's one of those heroes," said Susan Etlinger, an AI expert at the Altimeter Group. "She's someone who has, at great cost to herself, persisted in identifying, publicizing and trying to remediate a lot of the issues that arise with the use of intelligent technologies."

A number of people on her own team and others within Google tweeted their support for Gebru and anger with their employer. That includes Alex Hanna, a senior research scientist on the ethical AI team, who said that "to call her unbecoming of a manager is the height of disrespect." Dylan Baker, another Ethical AI team member, called her "the best manager I've had."

Gebru's departure could be damaging for Google's reputation in the ethical AI community and among tech workers broadly. The support for Gebru in the industry is nearly unanimous, and every leader who spoke to Protocol for this story echoed the same two sentiments: She is among the best at her technical work and Google's decision to fire her shocks and angers them. "The idea that this is going to be able to happen, and it's going to go away and it's not going to have an impact on tech … Google really needs to really look at itself in a mirror," Nkonde said.

All of the industry leaders who spoke with Protocol voiced their fear that her firing would have a chilling effect on other ethical researchers in the industry and at Google specifically. Academics and activists have long expressed skepticism about the integrity of ethical AI research at places like Google, but Gebru's reputation and leadership role lent credibility to Google's research and helped quell the critics. Earlier this year, Google even announced plans to launch an ethical AI consultancy that would provide tips for difficult problems learned from Google's own research and experience.

In firing her, Google not only gave up the voice that earned the ethical AI team respect in the first place, but also made it clear that there were consequences for speaking up, said Ansgar Koene, the global AI ethics and regulatory leader at EY and senior research fellow at the University of Nottingham. "Their division does great work, except a lot of the times they have their hands tied behind their backs because of such repressive policies," said Abhishek Gupta, a machine-learning engineer at Microsoft and founder of the Montreal AI Ethics Institute.

Gebru's firing was not entirely unexpected for people who knew her, including Gupta. Just the day before, while Gebru battled to get approval for the ethics research paper, Gupta and Gebru discussed how to create a legal system of protection for ethics whistleblowers inside tech companies. A few days before that, Gebru tweeted publicly that she wished there were a system of whistleblower protections.

"In a sense, this has been a long time in the making. This has, in bits and pieces, happened in the past, where she's tried to bring up relevant issues, and Google has sort of tried to suppress what she's saying," Gupta said, adding: "It's an unfortunate combination of what has been going on for months, I think."

Moving forward, people in her position need significant legal support to be able to express their concerns without fear of losing their jobs, said David Ryan Polgar, the founder and executive director of All Tech is Human. "There's a chilling effect for the people who don't have any type of national stature … You should have the ability to be a roadblock to what you would deem inappropriate activity."

And beyond the research work itself, firing Gebru makes Black women like her less likely to pursue the same career path, AI for the People's Nkonde said. "As Black women in tech, we all face similar issues, and not everybody is going to take the stand to stay within [the] industry," she said. For research scientists currently in school, choosing to work in the industry is far more intimidating after watching Gebru's experience play out, a feeling expressed by a number of those students on Twitter today.

If Gebru had decided to leave Google and announced that she would be going elsewhere, the reaction would have been celebratory, Nkonde explained. Instead, Google's decision to not only fire her but directly email the team she had managed about her departure creates a sense of fear and anger, showing that the tech sector, and Google specifically, "can be a hostile place for Black women," Nkonde said.

Ellen Pao, co-founder and CEO of Project Include and former CEO at Reddit, said that by firing Gebru, Google created an unfixable PR problem that illustrates a more systemic discrimination problem. "When I see Google in the context of its past, it has a terrible record of dealing with bias and discrimination, and it has a record of not hiring people from marginalized, underrepresented groups, not promoting them," she told Protocol.

"I think what it says is actually more important than what it says about Google. What this says is that the work of trying to remediate bias and create fairer technical systems is incredibly hard, and it's not just hard from a computational perspective. It's not just hard from a technical perspective. It's hard because it requires diversity of perspective, it requires diversity across many axes," Altimeter Group's Etlinger said.

Issie Lapowsky contributed additional reporting.

 Anna Kramer

Anna Kramer is a reporter at Protocol (Twitter: @ anna_c_kramer, email: akramer@protocol.com), where she writes about labor and workplace issues. Prior to joining the team, she covered tech and small business for the San Francisco Chronicle and privacy for Bloomberg Law. She is a recent graduate of Brown University, where she studied International Relations and Arabic and wrote her senior thesis about surveillance tools and technological development in the Middle East.

artificial intelligence ethics google workers activism
health tech
Protocol | China

Everything you need to know about the Yuanxin Technology IPO

Image: Yuanxin

Yuanxin Technology doesn't hide its ambition. In the first line of its prospectus, the company says its mission is to be the "first choice for patients' healthcare and medication needs in China." But the road to winning the crowded China health tech race is a long one for this Tencent- and Sequoia-backed startup, even with a recent valuation of $4 billion, according to Chinese publication Lieyunwang. Here's everything you need to know about Yuanxin Technology's forthcoming IPO on the Hong Kong Stock Exchange.

What does Yuanxin do?

There are many ways startups can crack open the health care market in China, and Yuanxin has focused on one: prescription drugs. According to its prospectus, sales of prescription drugs outside hospitals account for only 23% of the total healthcare market in China, whereas that number is 70.2% in the United States.

Yuanxin started with physical stores. Since 2015, it has opened 217 pharmacies immediately outside Chinese hospitals. "A pharmacy has to be on the main road where a patient exits the hospital. It needs to be highly accessible," Yuanxin founder He Tao told Chinese media in August. Then, patients are encouraged to refill their prescriptions on Yuanxin's online platforms and to follow up with telehealth services instead of returning to a hospital.

From there, Yuanxin has built a large product portfolio that offers online doctor visits, pharmacies and private insurance plans. It also works with enterprise clients, designing office automation and prescription management systems for hospitals and selling digital ads for big pharma.

Yuanxin's Financials

Yuanxin's annual revenues have been steadily growing from $127 million in 2018 to $365 million in 2019 and $561 million in 2020. In each of those three years, over 97% of revenue came from "out-of-hospital comprehensive patient services," which include the company's physical pharmacies and telehealth services. More specifically, approximately 83% of its retail sales derived from prescription drugs.

But the company hasn't made a profit. Yuanxin's annual losses grew from $17 million in 2018 to $26 million in 2019 and $48 million in 2020. The losses are moderate considering the ever-growing revenues, but cast doubt on whether the company can become profitable any time soon. Apart from the cost of drug supplies, the biggest spend is marketing and sales.

What's next for Yuanxin

There are still abundant opportunities in the prescription drug market. In 2020, China's National Medical Products Administration started to explore lifting the ban on selling prescription drugs online. Although it's unclear when the change will take place, it looks like more purely-online platforms will be able to write prescriptions in the future. With its established market presence, Yuanxin is likely one of the players that can benefit greatly from such a policy change.

The enterprise and health insurance businesses of Yuanxin are still fairly small (accounting for less than 3% of annual revenue), but this is where the company sees an opportunity for future growth. Yuanxin is particularly hoping to power its growth with data and artificial intelligence. It boasts a database of 14 million prescriptions accumulated over years, and the company says the data can be used in many ways: designing private insurance plans, training doctors and offering chronic disease management services. The company says it currently employs 509 people on its R&D team, including 437 software engineers and 22 data engineers and scientists.

What Could Go Wrong?

The COVID-19 pandemic has helped sell the story of digital health care, but Yuanxin isn't the only company benefiting from this opportunity. 2020 has seen a slew of Chinese health tech companies rise. They either completed their IPO process before Yuanxin (like JD, Alibaba and Ping An's healthcare subsidiaries) or are close to it (WeDoctor and DXY). In this crowded sector, Yuanxin faces competition from both companies with Big Tech parent companies behind them and startups that have their own specialized advantages.

Like each of its competitors, Yuanxin needs to be careful with how it processes patient data — some of the most sensitive personal data online. Recent Chinese legislation around personal data has made it clear that it will be increasingly difficult to monetize user data. In the prospectus, Yuanxin elaborately explained how it anonymizes data and prevents data from being leaked or hacked, but it also admitted that it cannot foresee what future policies will be introduced.

Who Gets Rich

  • Yuanxin's founder and CEO He Tao and SVP He Weizhuang own 29.82% of the company's shares through a jointly controlled company. (It's unclear whether He Tao and He Weizhuang are related.)
  • Tencent owns 19.55% of the shares.
  • Sequoia owns 16.21% of the shares.
  • Other major investors include Qiming, Starquest Capital and Kunling, which respectively own 7.12%, 6.51% and 5.32% of the shares.

What People Are Saying

  • "The demands of patients, hospitals, insurance companies, pharmacies and pharmaceutical companies are all different. How to meet each individual demand and find a core profit model is the key to Yuanxin Technology's future growth." — Xu Yuchen, insurance industry analyst and member of China Association of Actuaries, in Chinese publication Lanjinger.
  • "The window of opportunity caused by the pandemic, as well as the high valuations of those companies that have gone public, brings hope to other medical services companies…[But] the window of opportunity is closing and the potential of Internet healthcare is yet to be explored with new ideas. Therefore, traditional, asset-heavy healthcare companies need to take this opportunity and go public as soon as possible." —Wang Hang, founder and CEO of online healthcare platform Haodf, in state media China.com.

Zeyi Yang
Zeyi Yang is a reporter with Protocol | China. Previously, he worked as a reporting fellow for the digital magazine Rest of World, covering the intersection of technology and culture in China and neighboring countries. He has also contributed to the South China Morning Post, Nikkei Asia, Columbia Journalism Review, among other publications. In his spare time, Zeyi co-founded a Mandarin podcast that tells LGBTQ stories in China. He has been playing Pokemon for 14 years and has a weird favorite pick.
health tech healthcare insurance
sponsored content
Sponsored Content

From Home to Hybrid: Asana’s COO Shares Back-to-Office Advice on Productivity

The way we work has fundamentally changed. COVID-19 upended business dealings and office work processes, putting into hyperdrive a move towards digital collaboration platforms that allow teams to streamline processes and communicate from anywhere. According to the International Data Corporation, the revenue for worldwide collaboration applications increased 32.9 percent from 2019 to 2020, reaching $22.6 billion; it's expected to become a $50.7 billion industry by 2025.

"While consumers and early adopter businesses had widely embraced collaborative applications prior to the pandemic, the market saw five years' worth of new users in the first six months of 2020," said Wayne Kurtzman, research director of social and collaboration at IDC. "This has cemented collaboration, at least to some extent, for every business, large and small."

Keep Reading Show less
Kate Silver

Kate Silver is an award-winning reporter and editor with 15-plus years of journalism experience. Based in Chicago, she specializes in feature and business reporting. Kate's reporting has appeared in the Washington Post, The Chicago Tribune, The Atlantic's CityLab, Atlas Obscura, The Telegraph and many other outlets.

sponsored content sponsored
Protocol | Workplace

How to make remote work work

Hofy made an early bet that COVID-19 would have a long-term impact on workplaces. The company recently raised $15.2 million for its remote workforce equipment management solution.

Hofy recently raised $15.2 million for its remote workforce equipment management service.

Photo: Jannis Brandt/Unsplash

It's your new employee's first day of remote work, but their laptop hasn't shown up yet. Not a good look.

This very 2021 persistent problem is part of why Hofy, a remote workplace management tool, recently raised $15.2 million to help companies deploy laptops, chairs, desks and other physical equipment to their remote employees. The idea for Hofy, which is launching out of stealth today, emerged in the early days of the COVID-19 pandemic — before lockdowns went into effect in the U.S. and the U.K. Hofy's co-founders, Sami Bouremoum and Michael Ginzo, had a feeling that COVID-19 would have a long-term effect on society.

Keep Reading Show less
Megan Rose Dickey

Megan Rose Dickey is a senior reporter at Protocol covering labor and diversity in tech. Prior to joining Protocol, she was a senior reporter at TechCrunch and a reporter at Business Insider.

remote work
Protocol | Policy

Tech giants want to hire Afghan refugees. The system’s in the way.

Amazon, Facebook and Uber have all committed to hiring and training Afghan evacuees. But executing on that promise is another story.

"They're authorized to work, but their authorization has an expiration date."

Photo: Andrew Caballero-Reynolds/AFP via Getty Images

Late last month, Amazon, Facebook and Uber joined dozens of other companies in publicly committing to hire and train some of the 95,000 Afghan refugees who are expected to be resettled in the United States over the next year, about half of whom are already here.

But nearly two months since U.S. evacuations from Kabul ended and one month since the companies' public commitments, efforts to follow through with those promised jobs remain stalled. That, experts say, is partly to do with the fact that the vast majority of Afghan arrivals are still being held at military bases, partly to do with their legal classification and partly to do with a refugee resettlement system that was systematically dismantled by the Trump administration.

Keep Reading Show less
Issie Lapowsky

Issie Lapowsky ( @issielapowsky) is Protocol's chief correspondent, covering the intersection of technology, politics, and national affairs. She also oversees Protocol's fellowship program. Previously, she was a senior writer at Wired, where she covered the 2016 election and the Facebook beat in its aftermath. Prior to that, Issie worked as a staff writer for Inc. magazine, writing about small business and entrepreneurship. She has also worked as an on-air contributor for CBS News and taught a graduate-level course at New York University's Center for Publishing on how tech giants have affected publishing.

afghanistan big tech hiring refugees trump
n26
Protocol | Fintech

How European fintech startup N26 is preparing for U.S. regulations

"There's a lot more scrutiny being placed on fintech. We are definitely mindful of it."

In an interview with Protocol, Stephanie Balint, N26's U.S. general manager, discussed the company's approach to regulations in the U.S.

Photo: N26

N26's monster $900 million funding round announced Monday underlined the German startup's momentum in the digital banking market.

Stephanie Balint, N26's U.S. general manager, said the funding will be used for expansion and also to improve "our core offering to make this the most reliable bank that our customers can trust," she told Protocol.

Keep Reading Show less
Benjamin Pimentel

Benjamin Pimentel ( @benpimentel) covers fintech from San Francisco. He has reported on many of the biggest tech stories over the past 20 years for the San Francisco Chronicle, Dow Jones MarketWatch and Business Insider, from the dot-com crash, the rise of cloud computing, social networking and AI to the impact of the Great Recession and the COVID crisis on Silicon Valley and beyond. He can be reached at bpimentel@protocol.com or via Signal at (510)731-8429.

n26 neobank
Latest Stories