SOTN2023-02 Keynote with Alexander "amac" Macgillivray

7:57AM Mar 16, 2023

Speakers:

Alexander Macgillivray

Keywords:

ai

privacy

people

president biden

data

technologists

americans

automated systems

technology

cto

advance

ftc

policy

federal government

important

harms

lead

spectrum

protections

tech

Hey, folks, it's absolutely wonderful to be back here. Thank you very much, Shane, for that kind introduction. And thank you for state of the net for having me here today. It's not every day that I get to be around a collection of techie legal policy wonks all put together. And I feel like we're a rare breed. And it's nice to be among my type of geeks. As Shane mentioned, I lead the US chief technology officers office, or the CTO office. And I served in a similar role in the Obama Biden administration. But a lot has changed since then. I'll talk a little bit about that. And then about privacy, AI, broadband access, and the three things that the CTO shop is really doing a ton of work on. At the beginning of the Obama administration, there were a lot fewer technologists, in the Executive Office of the President, and those that were there were much more siloed. And often, people thought that there was a really bright line between tech policy issues like network neutrality being an example from the Obama administration, and then the rest of policy, like health care, and that those two things were entirely separate and distinct from each other. But because of some hard earned experience, and the work of so many folks in this room, there's really now been a pretty big shift in how we think about all that. Today, people across the government understand that in the modern age, all technology is technology policy. Sorry, I think I said that wrong. All technology is technology policy. No all policy is technology policy. Today, it's abundantly clear why technologists are important in shaping policy and its implementation. today. We know that technologists, the technologists required to develop sound government policy can take many forms. From data scientists like the US chief data scientist and US Deputy CTO Denise Ross, to computer scientists, tech focus lawyers like me, and social scientists. In the Biden, Harris administration, there are many technologists across the White House and in senior positions across federal agencies, the US digital service, the General Services Administration's tech transformation service, the Federal CIO and agency CIOs, the digital services teams, all of these folks are out there as federal techies and are focused on making government services better and more modern. So we've made a good start. We haven't gotten to a government where there's as many values led technologists as there are economists and lawyers. We're still outnumbered at most meetings. But we are moving in the right direction. Last September, the White House released its principles for enhancing competition, and tech platform accountability. And in January, the President called on Congress to come together and pass bipartisan legislation to hold big tech accountable. In an op ed in the Wall Street Journal. As the President said, we still have much work to do to be able to finish the job. For us in the CTO office, that job means pushing forward responsible innovation, where innovation doesn't just happen for innovation sake, but as a way of purposely advancing core American values, such as opportunity, privacy, and equity. where technology is aligned with and bolsters democratic principles, and human rights. There was something that DJ Patil, the very first US chief data scientist, had a way of saying that I think still rings very true in this administration, which is our job is to move purposefully and fix things, not move fast and break things. We need innovation that benefits all Americans technology that uplifts all people strengthens democracy. Safeguards civil rights, includes everyone and brings prosperity really widely. We can advance these values by pursuing two things purposeful technology, technological innovation, and strong protections for Americans rights and safety. And privacy is an important foundation. At the State of the Union, President Biden made clear that we need serious federal protections for Americans privacy. Presidents push for strong privacy legislation comes against the backdrop of a litany of privacy horror stories, I'll just give you a few. And just within the health privacy field, ad targeting platforms have been selling the targeting of people with clinical depression. apps have been leaking people's health information to other companies without users permission, or even their knowledge. data brokers have been selling mobile IDs tied with people who are actively pregnant, which is especially worrisome given the attacks on reproductive rights that we are seeing. So the problem is clear. And so as the urgent need for action, we all know that we need limits on the data that companies can collect us and share. That means limits on the collection and use of data related to your Internet history. Your personal communications, your location, your health, your genetic and biometric data. It's not enough for companies to disclose what data they're collecting. Much of that data shouldn't be collected at all. And these protections need to be especially strong for children who are particularly vulnerable to harm. Unchecked data collection by companies can also end up in the hands of our adversaries and competitors, and be exploited in ways that undermine our economic and national security. We're taking big steps to address these problems. President urged Congress to pass strong federal privacy legislation, the United States and 60 partners around the world made clear that individuals should have their privacy protected online, in the declaration for the future of the Internet. And the White House elevated privacy as a central principle and the blueprint for an AI Bill of Rights, which I'm going to talk a little bit more about later. With us prize challenges to push this research forward, and research strategies to encourage the development of privacy enhancing technologies, kind of have your cake and eat it too. technology that would allow people and society to benefit from the use of data while maintaining privacy. And we've recruited a leading expert on privacy and tech policy more broadly, who's sitting right here in the audience, Professor Deidre Mulligan, who I can't, I'm really excited to be working with to help lead the government work as a US Deputy CTO. Other parts of the federal government are also taking important steps on privacy. The NTIA I know Alan will be up here soon, is leading a comprehensive review of the nexus between Privacy Policy and Innovation in the Internet economy. They are currently seeking public input through a request for common and listening sessions on the intersection between privacy civil rights and the digital economy. Earlier this year, they also released an important report highlighting the need for policy interventions to approve improve privacy in mobile app in the mobile app ecosystem while supporting strong competition in that ecosystem. The FTC has also shown leadership in protecting Americans privacy and civil rights against harms from excessive data collection, the use of automated systems and defaults in the gaming platforms that allow unknown adults to digitally interact with kids. They announced an advance notice of proposed rulemaking to crack down on commercial surveillance, and the data gathering practices that underpin it. And they recently stood up a new office of technology that will continue advancing the FTC has technical expertise and capacity to protect Americans privacy, because we need more people with that technological background to be able to do good enforcement in this modern age, and they're hiring. And if you're a technologist, who has always been a little bit FTC, curious, now is definitely the time to take the plunge. And to go help. If you know somebody who fits that bill, please do send them the FTC is way it's a really important office. That is that is as I say building. We're also working to make sure that the data that is collected is us that is used in ways that are fair and equitable. In fact, we are finding ways to use that data to advance equity. That means putting in place new practices to understand data such as people's race and gender, and then use that data to identify and fix disparities in federal programs. To answer the question of how federal services are delivered in terms of that type of data. We're also studying how police, State Tribal, Local and Territorial law enforcement agencies across the country are handling data about police activities. In order to help inform more fair and just policing. We've seen a bunch of the good ways that data is used to help make people's lives easier. But no one in this room is naive to the fact that data is often used to power AI systems that are used in concerning ways. There's lots of talk right now about sentient AI or or some sort of future robot doomsday scenarios involving paperclips.

But the truth is that other harms, to our safety, our rights, our opportunities, and our access are happening right now. Today, women's job applications are being unjustly rejected by AI. Students are being falsely accused of cheating by automated systems, black Americans are being denied life saving health care by algorithms. We can't afford to wait to mitigate algorithmic harms and discrimination. All Americans should fully benefit from AI systems that strengthen democracy, advanced equity and create a more secure and prosperous world. And President Biden has stated that we need more transparency into the algorithms to be able to stop them from being used to discriminate against Americans. Last month, President Biden signed an executive order, sorry, and he has made clear that this is not just an aspiration, it's the policy of the United States government. And that's the executive order that he signed last month that directs the federal government to procure and use automated systems in a manner that advances equity, not just hold it steady but advances it. He also directed federal agencies to affirmatively advance civil rights By seeking opportunities to protect people from algorithmic discrimination, we've laid out a way to get that directive done. Last fall, we published principles and practices that can make sure that AI systems live up to these important aspirations. Our office with Dr. Alondra Nelson science and society division led the development of the blueprint for an AI Bill of Rights. This document lays out the rights that people should have to protect us from the harms of automated systems. It's an articulation that Americans should expect better, and demand better from their technologies. In short, automated systems like aI should include at least five core protections. One, they should be safe and effective. They should protect against algorithmic discrimination, they should respect privacy, they should ensure notice and explanation, and they should establish human alternatives to AI. We've all been in that system where we can't get the thing done that we need to get done. Because of an automated system, there needs to be a way to talk to a human. The blueprint for an AI Bill of Rights provides concrete practices to achieve these goals. It's a resource for everyone, from technologists building AI systems, to policymakers that set the rules of the road to advocates, challenging institutions to do better as they use automated systems. Alongside the brute blueprint, we announced a slate of actions from government agencies, from the Department of Housing and Urban Development, to education, to health and human services, to Veterans Affairs, to advance these principles. And we are hard at work to continuing to implement these protections. The exciting thing about all this is I really do believe that AI is among the most truly important technologies of our generation, right up there with the Internet, where I spent most of my career to the cell phone which is ever present. It is also developing extraordinarily quickly, due to improvements in compute, and the data used for creating the models. The promise is huge, but so are the potential risks. And it's going to take all of us using all of our tools in order to make sure we get it right. One of the reasons I'm proud to be at the CTO team is that we are the home of the congressionally mandated national AI initiative office, which we pronounce niho Kind of like California knio knio helps coordinate key national AI priorities. It's developing a forthcoming update to the 2019 ai r&d strategic plan to support research and development that can better align AI with the public good. We just published an implementation plan for the National AI research resource that would provide America's researchers with the tools and data needed to develop cutting edge AI systems, the kind of stuff that is today too often only available to well resourced tech companies. Increasing who is able to do AI research and development will increase America's competitiveness in AI. But it's also critical to bring more people to the table in researching this important tool. This work complements the critical work of the National Institute of Standards and Technology NIST, is doing to ensure that AI systems work for the American people. Earlier this year, NIST released its AI risk management framework, which is a vital step toward mitigating the full suite of risks posed by AI systems. And it builds on the guidance that the FTC has put out and reaffirmed just a few days ago, to help businesses utilize AI in a manner that follows consumer protection laws and regulation. And that brings me to my second job pitch. The federal government needs more AI expert. We need technical experts, socio technical experts, even lawyers and policy wonks, one benefit of working for the federal government as opposed to say, being a journalist, like Kevin ruse is that no chat bot has so far tried to steal the wife of a of a federal government employee. And I'm not saying that they'll never come for us, but I really hope that they do not. But seriously how we as a country establish strong boundaries for AI while pushing hard to realize its many promises. The possibilities the President Biden speaks so eloquently about is hugely important. And working for the government is a rewarding way to contribute to that work. Finally, I want to talk a little bit about broadband. I think I saw Alan walk in so I'm not going to take too long, but much of the benefits of technology are still unavailable to many Americans. As President Biden has said, high speed Internet is not a luxury any longer. It's a necessity. But for millions of Americans, without an affordable, reliable broadband connection, those benefits have been out of reach for too long. In some places, broadband is simply not available. In others. It's the price that puts broadband out of reach. Across the Biden and Harris administration. We're making investment in closing the infrastructure gaps and making high speed Internet more affordable. Those efforts include the affordable connectivity program that provides a $30 credit On a household Internet bill, and for many families that'll bring the cost down to $0. For broadband, you may qualify or know someone who does so please do spread the word, the URL is Git internet.gov. It is a useful program and still ongoing and picking up people. Furthermore, NTIA, USDA and Treasury are all funding infrastructure investments that will fill gaps in our Internet infrastructure and make high speed reliable connections available in every community. We're working to make sure those investments are equitably distributed, and that they reach communities, especially tribal and rural communities that are often left out. We lead interagency efforts to connect rural and tribal lands work that is overcoming barriers to access to funding, improving data accuracy, and promoting digital inclusion. We're also helping shape our nation's wireless future through innovative spectrum policy. Practically every part of the vision President Biden is laid out for our country has a spectrum dependent component. You all people in this room are some of the few that really know how true that statement is. We want to lower costs for American families and increase equity for low income people who are more likely to rely on their smartphones to get online. We want telemedicine that meets people where they are. We want cutting edge research to fight climate change and advanced forecasting to protect Americans during severe weather. We want national defense systems that are ready when we need them. These priorities and many, many more all require spectrum resources. If you're interested in spectrum or broadband I know many of you are there are plenty of great jobs in the federal government of course, I'd also say particularly to if you're in high school or college and maybe watching this online, or if you know a high school and college person who is who is technically minded. Have them think about spectrum engineering. We really desperately need more spectrum engineering folks in the federal government. And it's, it's a really important part of how we get a lot of our job done. So on that note, I'm going to wrap up before I start pitching cybersecurity jobs. Thank you again to Tim Lordan in the state of the NET team for having me. I look forward to working with you all to make technology work for all Americans and to advance technology for democracy, both at home and abroad. Thank you very much