Google Rehauls AI Ethics Research Staff Following Accusations Of Racism
By Sabrina Lynch
- Google plans to increase its AI team to 200 researchers to avert race discrimination in tech
- The tech giant fired its second AI Ethics Head over diversity disputes
Tech giant Google (GOOG +2.05%) plans to double its Artificial Intelligence (AI) research team to 200 employees, after ousting two female AI experts, Dr. Timnit Gebu, a Black woman, and Margaret Mitchell. The company demanded Dr. Gebu retract a paper that criticized an AI technology that powered some of Google’s search results. She was part of the company’s 1.2% Black female employees, and was delegated the responsibility of ensuring its AI tech was implemented without causing harm to communities of color. Despite a pledge to commission an investigation into the circumstances around the AI experts’ departures, Google has a lot of bridges to build to restore trust in its commitment to produce ethical technology that does not encourage systematic racism.
Why This Matters: Leaders, influencers and academics within the tech space have been calling for greater scrutiny of emerging digital technologies that can be used to uphold racial inequality, discrimination and intolerance. This is why many people were shocked that when Dr. Gebu called Google out, she was fired, contrary to its statement claiming they had accepted her resignation. To add fuel to the fire, a Google engineer recently came forward claiming Gebu and Mitchell were systematically frozen out.
The chaos in Google’s AI leadership fractured the company’s relationship with external groups who focus on increasing the presence of Black people in the field of Artificial Intelligence. Black in AI released a joint statement with Queer in AI and Widening NLP criticizing Google for setting a “dangerous precedent for what type of research, advocacy, and retaliation is permissible in our community”.
In 2018, researcher Joy Buolamwini at the Massachusetts Institute of Technology discovered three facial analysis programs from major tech companies, including IBM (IBM +0.48%) and Microsoft (MSFT +1.38%), carried gender, skin-type and age biases. When their programs read photos of lighter-skinned men, they misidentified sex approx 1% of the time. However, the darker the skin in the photo, the larger the error rate became.
This study enraged critics of AI, driving a backlash against facial recognition tech and its use in law enforcement. Specifically, 44 civil rights groups wrote an open letter to Amazon insisting they immediately stop selling facial recognition technology to the Police Force, pointing to numerous examples of police violence against the Black community.
Situational Awareness: Ethical AI has become the new darling of the tech world. Its usage can either diminish or exacerbate racial equality. Google faces a long road to salvage its reputation as a brand that makes diversity, equity and inclusion part of how they build their products. As the newest technology that’s ripe for innovation, AI will be a crucial tool to help progress social equity.
CBx Vibe: “Blow the Whistle” Too $hort