CHANGEMAKERS: Renée Cummings on AI, Criminal Justice, and the Whiteness of Silicon Valley

Renée Cummings is a criminologist, criminal psychologist, AI ethicist, and AI strategist. As the founder of Urban AI, a community scholar at Columbia University, and a founding board member of Springer’s new AI and Ethics Journal, she is on the frontline of ethical AI, battling bias, discrimination, and injustice and advocating for more diverse, equitable, and inclusive artificial intelligence. Renée specializes in AI for social good, justice-oriented AI design, social justice in AI policy and governance, and using AI to save lives. Renée spoke with All Tech is Human’s Andrew Sears about AI’s potential to transform urban life and the challenges that are being encountered along the way.

Andrew: You come to your work in AI ethics from a background in criminal justice. How does your experience in criminal justice inform how you approach the field of AI ethics?

Renée: I am extremely passionate about the power, pervasiveness, and potential of AI to design and deploy real time evidence-based solutions to many criminal justice and quality of life challenges. I believe in using AI to rewrite the policing playbook. As calls are amplified to defund the police and radically reimagine policing, I believe the criminal justice system will lean on AI to expedite predictive policing strategies to appease the police and the public. 

We have got to be data conscious and data vigilant because without the requisite checks and balances, throughout the design, development and deployment life cycles, AI could be used to deliver very punitive and oppressive strategies that could undermine civil liberties, civil rights, community cohesion and community resilience. We have already experienced some of the perils of facial recognition technologies, predictive policing, risk assessment tools, for sentencing, that overestimate recidivism and frustrate due process. We’ve seen the risks of outsourcing criminal justice decisions to algorithms. 

Therefore, as I am committed to innovation, I am also committed to ethical design and a social justice approach to AI, ensuring justice-oriented design is a prerequisite for crime control and crime prevention. I believe criminologists have a critical role to play in co-designing the future and I continue to say criminology must be the conscience of AI.

You’re the founder and CEO of Urban AI. What unique challenges does AI present for urban communities, and what sorts of solutions are you working towards?

Cities are the laboratories of now and the future. AI is an absolute game changer, a futuristic Bunsen burner. As we negotiate our way out of this pandemic and design data-driven strategies to prevent future pandemics, the concept of Urban AI will flourish. We will seek urban AI solutions for more contactless interfaces, telehealth, telelearning, more connected communication and intelligent transport systems, food security, climate security, public safety, and an improved quality of life. With Urban AI, I am combining epidemiological criminology, urban criminology, urban technology, and urban culture to ensure equity in the access to resources, inclusive innovation, diverse, principled, and ethical design of AI systems. Now, I just need some investors!   

There’s been a spotlight on AI surveillance technology recently, particularly on how it can contribute to racially biased policing. What other implementations of AI technologies do you see contributing to systemic racism?   

Systemic racism is so deeply rooted in everyday life that no application of AI is immune to its debilitating effects. Therefore, every implementation of AI must be conscious of systemic racism, bias, and discrimination and ensure there are robust detection and mitigation strategies as well as inclusive design which respects and celebrates diversity and equity. AI is already doing some reflection, with some tech companies abandoning further development of facial recognition technologies that discriminate and committing to advancing racial justice, according to their press statements, either by hiring, investing in, or supporting black talent. While racial justice may now be de rigueur for tech companies, what is required is an end to corporate tokenism and a commitment to meaningful and sustained organizational change. Not a PR exercise, but a new consciousness in doing business. 

Much has been written about the whiteness of Silicon Valley; how the people who are designing and building our technological future are predominantly white. What are some of the blindspots that white technologists tend to have when envisioning the future, and what unique insights do people of color bring to conversations about the future?

Silicon Valley needs to do some deep soul-searching. Tech companies, over the years, have expended extraordinary resources on diversity, inclusion, anti-discrimination and implicit bias training and best practice organizational policies. Have the efforts and investment amounted to little more than window dressing? Are you going to design an algorithm to teach you how to be human and have compassion, empathy, unconditional positive regard and be committed to equality and justice? Twenty years from now, or less, the majority of the U.S. population will be people of color. Therefore, every conversation about the future is a conversation about people of color.

You’re a founding board member for Springer’s new AI and Ethics Journal. What are some of the unresolved questions on the frontier of AI ethics that you’re hoping will be explored in this journal?

This is an opportunity to participate in an exercise that will future proof our tomorrow, today. This is important to me because I am committed to ethical AI. I am also committed to ensuring due process, duty of care and public engagement are perquisites for ethical design and inclusive innovation. The long-term impacts of technology on society must be a priority as we innovate. Protecting the public interest means providing a forum for researchers, developers, users, public policy makers, and the public to better understand the challenges of ethics in AI, and to address them in ways that don’t stymie innovation but encourage due diligence, diversity, and the understanding that vulnerable groups must be visible in AI and not revictimized or further marginalized. Once we get that right, we would have answered many of the unresolved questions on the frontier of AI ethics.

You can keep up with Renée on LinkedIn and Twitter.

WATCH Renee Cummings on All Tech Is Human’s livestream for Covid-19, AI Ethics, & Surveillance

LISTEN to Renee Cummings’ podcast interview on Funny as Tech

Andrew Sears is an advisor at All Tech is Human and the founder of technovirtuism.