REMARKS OF COMMISSIONER GEOFFREY STARKS AT THE INCOMPAS 2023 POLICY SUMMIT WASHINGTON, D.C. Thank you, Angie, and good afternoon, everyone! It’s an honor to join you at this year’s Policy Summit, and to speak before professionals that have dedicated their careers to building competitive markets. We are living in a remarkable time in telecom history. From COMPTEL to INCOMPAS, your advocacy has been indispensable and has played a central role, over a number of years, in getting us to where we are. From moving on from the Baby Bell era, to securing opportunities for interconnection, to modernizing regulation for an era of IP and broadband. The sweep of telecom history has brought us to this moment for connectivity’s new age, defined by the twin goals of access and equity. These goals are not unfamiliar, but they feel imbued with new promise and fresh opportunity. A moment to be seized. That’s why I wanted to use my time to talk about why I feel the fierce urgency of now. I. First, we’ve known for years that in a networked world, competitive markets only work if they work for everybody. Certainly Congress and the FCC understood that when they acted to open our telecom markets back in the 1990s. They realized that although connectivity to one user unlocked value for every other user, competition alone wouldn’t drive new networks to rural homes, low-income households, and other vulnerable communities. They responded by creating universal service programs to ensure that a newly dynamic telecom market was also an equitable one. Those programs made substantial progress in the years that followed and continue to have an enormous impact today. But in a game of telecom cat and mouse, innovation in connected technologies also progressed and expanded our baseline connectivity requirements. Healthcare, education, entertainment, employment—virtually every aspect of American life embraced broadband and bandwidth. At the same time, wireless climbed to higher speeds, and from seamless roaming to near-nationwide coverage. Together with GPS, WiFi, touchscreens, and processing power, it fueled a smartphone, tablet, and IoT revolution, ushering us into a new era of connectivity for everything and everywhere, including within our own homes. Before long—and certainly by 2020—these changes made clear that we had significant challenges when it came to equity. We had many more communities left to connect and serious adoption barriers still to breakdown—chief among them, affordability. The moral and economic case for ubiquitous, high-quality access had not only become stronger, it had transformed into a fundamental matter of civil rights and American competitiveness. The fierce urgency of now, for everybody. A bipartisan Congress responded with force in the Infrastructure Investment and Jobs Act, meeting the moment with a once-in-a-generation commitment to connecting every American household with high-speed internet. The centerpiece of the law’s broadband effort lies in BEAD and the Affordable Connectivity Program, or ACP. BEAD put $42.5 billion into building new networks that reach primarily unserved and then underserved homes—and thereby addresses gaps in our broadband supply. ACP put $14.2 billion into lowering the cost of broadband service for vulnerable households—and thereby ensures that economic insecurity does not depress our broadband demand. The two programs work as a pair. As it turns out, building a network doesn’t strictly solve for the millions of Americans who struggle to simply afford broadband service. At the same time, deploying new networks becomes less risky and less expensive when more of the serviceable population can actually afford the end product. In that way, ACP enables BEAD to reach more homes. To get broadband to everybody and everywhere, supply and demand both matter, and they’re both directly related. This, along with the urgent need for ubiquitous connectivity, were the key insights behind the infrastructure law. That brings us to 2023. Now, in light of the IIJA’s enactment, we have turned to implementation. State and local leaders, members of Congress from both parties, and the President and Vice President have urged that we move with all deliberate speed to reach our shared objective of internet for all. I agree—and that’s why I believe it is essential to sustain funding for the Affordable Connectivity Program. With enrollment picking up, we’ll need to address the longevity of the program’s funding before too long. We’re much better off addressing the issue quickly and with certainty—and there are three basic reasons why that is so. First, we must keep the millions of Americans who have signed up for the program secure in their access to broadband. As of this week, 16.6 million households rely on ACP for a broadband connection. They live everywhere in the country. All fifty states have at least 10,000 households enrolled. 33 states have at least 100,000 households enrolled and with enrollment in the 90,000s, West Virginia and my home state of Kansas are right there knocking on the door. Rural communities in particular have a strong foothold in this program; in fact, several thousand rural zip codes, and more than 80% of non-metro counties, have at least one hundred ACP households signed up. We can all agree that these families have enough to worry about. Losing the connectivity they need to succeed, and to help their children succeed, should not be on their already long list. In fact, the case for keeping them connected has only become stronger. Telehealth adoption continues its impressive growth—and as we work to make telehealth even better and safer, we’re also noticing connectivity-driven disparities. In particular, low-income families have lower rates of video telehealth use, due in part to lower rates of broadband adoption. Madjid Karimi, Euny C. Lee et al., National Survey Trends in Telehealth Use in 2021: Disparities in Utilization and Audio vs. Video Services, Assistant Secretary for Planning and Evaluation, U.S. Dep’t of Health and Human Servs. (Feb. 2022). At the same time, the American economy continues to soar down a digital path and will soon approach a point where a substantial majority of all jobs, and the vast majority of high-earning jobs, will require advanced digital skills. These are skills that you build over time by using a connected device over a home broadband connection—the exact tools that ACP provides. Let’s be clear. Connectivity today is about more than just STEM or STEAM. It’s about equal access to opportunity, whatever you ultimately choose to do for a living. 16.6 million households connected is an incredible number. But I see even more momentum in this program. That brings me to the second reason why we need to fund ACP sustainably: there are millions more who are eligible, and we need to get them signed up. Since the day we rolled out ACP, I’ve been vocal about our need to push to reach every last one of the roughly 48 million households that are eligible for the program. Later this month, we’ll be announcing grantees for outreach programs that will raise awareness and boost eligible enrollment. These programs include Your Home, Your Internet, which will focus on signing up families on federal housing assistance—a vulnerable group of Americans with a particularly low rate of enrollment, and a pressing need for the help. But as our outreach efforts extend their reach, the last thing we need is the prospect of a future funding lapse looming over our enrollment efforts. I’ve traveled the country speaking to local officials and eligible residents about ACP and let me tell you, the potential impact here is substantial. Trust in government remains critical, and much of it revolves around the fear that the government could suddenly end the program. In fact, on several occasions, I’ve been asked specifically by folks in the community about what will happen to them if ACP funding stops. The problem will only grow worse as the cliff approaches. While it may not be affecting a significant number of sign-ups right at this moment, funding certainty will become an enormous enrollment barrier soon enough. It also bears emphasis that ACP secured participation from hundreds of providers, many of whom have committed to providing qualifying broadband free of charge with the subsidy. In fact, as the President and Vice President announced at the Rose Garden last year, providers covering more than 80 percent of Americans offer plans at qualifying speeds at no cost to the ACP consumer. We need that number to grow, not shrink, to get every eligible American connected, and certainty behind the program is key here. That leads me to the last reason, and it is equally important. We need to move quickly to secure ACP so that our BEAD dollars go as far as possible and succeed in connecting every unserved and underserved home. With BEAD, as NTIA Administrator Davidson has said, funding will begin flowing to states later this year. Before we know it, states will be defining projects and putting them out to bid and evaluating and choosing between competing proposals. That raises the question of how much funding each project will consume and how far those dollars will go in terms of connecting unserved and then underserved households. ACP enrollees tend to be good customers with low churn, and they add significantly to the subscriber base of the network. That means the availability of ACP directly affects the revenue side of the equation and therefore the size of the funding gap that states will have to bridge. In fact, a recent study quantified the impact of ACP’s availability on buildout in the rural areas that are BEAD’s focus. See Boston Consulting Group & Common Sense Media, Closing the Digital Divide Benefits Everyone, Not Just the Disconnected at 29-31 (Dec. 7, 2022). It concluded that ACP reduces the subsidy needed to incentivize a build in rural areas by a whopping 25 percent. Now that is a synergy. In the same way, ACP can improve the effectiveness of other broadband deployment programs. The IIJA in fact pushes BEAD provider-recipients to plan this way by requiring them to offer a low-cost broadband service option. So if we leave ACP’s future uncertain, we’re potentially reducing the impact of BEAD’s appropriation by as much as 25 percent. With an estimated 7.8 million unserved and about 6 million underserved locations based on current data, that could translate into potentially millions of American homes in rural areas remaining, unfortunately, on the wrong side of the digital divide. That weighs heavily on me, and it should for all of us. We can’t afford to leave behind the 16.6 million families that are already enrolled in this essential program. We can’t afford to undermine efforts to sign up the millions more who are eligible but unenrolled. We can’t afford to effectively raise the costs of BEAD in a way that could leave millions of rural homes unserved and underserved. And we certainly can’t afford to do all three. Putting ACP on a path toward sustainable funding will help us capitalize on this unprecedented moment and keep our promise to connect every American to high quality broadband. II. Before I wrap up, I’d like to talk about another fundamental change in how we view the issue of equity in connectivity’s new age. And that is the importance of not just protecting our data, but also making sure it isn’t used to make inequality worse. As many of you know, the Commission has long protected the privacy of telecom consumers, and even our first take on CPNI rules recognized the importance of putting the user in control. But the concern today is much more vast, because so much of our lives depends on connectivity. When we’re online, we read, we talk, we bank, watch, and learn. We search, we buy, we react, post, and comment. We do it with products we pay for, and we do it a lot with ones that are free. And each step of the way, we drop crumbs about who we are, where we live, what we do, how much we make, what we like, and what we believe. Those crumbs can then be used to serve ads, recommend content, make credit decisions, build products, and more. Sometimes, those crumbs can be kept around for no reason at all, waiting for a buyer or business case—and perhaps a cyberattack. Needless to say, we’re confronting new issues in the collection of data and how it’s used, including the potential for data-driven bias against communities that are vulnerable and underserved. I commend the NTIA for launching an inquiry into this very issue and was pleased to file comments just yesterday as an interested party. See Press Release, Federal Communications Commission, Commissioner Starks Comments on Privacy, Equity, and Civil Rights (rel. Mar. 6, 2023). Let me offer a few examples. For one, there’s the potential relationship between algorithmic bias and deployment. In the IIJA, Congress directed the Commission to promulgate rules to prevent and eliminate “digital discrimination of access based on income level, race, ethnicity, color, religion, or national origin.” We have a proceeding underway to carry out that responsibility, and commenters have raised concerns about discrimination driven by data collection and usage. Some have reminded us that providers use algorithms to determine where to deploy broadband and at what level of service. See, e.g., Comments of Free Press, GN Docket No. 22-69 at pp. 8-9 (filed May 16, 2022); Comments of The Utility Reform Network (TURN), GN Docket No. 22-69 at pp. 18-19 (filed May 16, 2022). These commenters have raised concerns that bias in those algorithms could lead to minority communities being disproportionately underserved. Similar concerns have been raised about the data underlying network upgrades and maintenance, broadband advertising, and subscription pricing models, and about the provision of less secure service to some communities. See, e.g., Comments of Lawyers’ Committee for Civil Rights Under Law, GN Docket No. 22-69, at pp. 18-19 (filed May 16, 2022); Comments of Leadership Conference on Civil and Human Rights, GN Docket No. 22-69 at p. 5 (filed May 16, 2022); Comments of Multicultural Media, Telecom and Internet Council, GN Docket No. 22-69, at p. 10 (filed May 16, 2022); Comments of National Urban League, GN Docket No. 22-69 at p. 4 (filed May 16, 2022). These concerns are important, and they warrant our careful consideration in the digital discrimination proceeding. Another example lies in the algorithms companies use to deliver captioning services to people who are deaf or hard of hearing. By statute, we’re responsible for ensuring that those who are deaf, hard of hearing, deafblind, or have speech disabilities can communicate in a manner that is functionally equivalent to those without such disabilities. To carry out that responsibility, we fund a program to give individuals access to IP-captioned telephone service, which allows them to listen to the person they’re calling while simultaneously reading captions of what the other person is saying. The program is vitally important to those who use it, and so is transcribing speech accurately and reliably. Those who rely on the service use it for just about everything—saying in touch with loved ones, talking to employers and prospective employers, and seeking help when they need it, including by dialing 988 and 911. Typically, an individual operator provides the captioning in real-time, but we’re seeing growing use of automated speech recognition, or ASR, to perform the captioning function. The question is whether ASR can do the job reliably for all people—which is a concern I’ve now raised in several proceedings. Studies have shown that speech recognition systems make far more errors when transcribing the speech of people of color, and we’ve seen similar bias in algorithms that power facial recognition. We need to closely examine whether these disparities affect the ASR used in our captioning programs. Finally, if you’ve heard me talk about the media industry lately, you’ll know that I’m excited about TV’s new IP-based transmission standard, ATSC 3.0. The potential benefits here are multiple, including more free, over-the-air content and higher quality picture and audio. But with new technology often comes new risk, and 3.0 is no exception. The transition will allow broadcasters to collect much more viewer data than they do right now and give them the opportunity to deliver targeted advertising. We need to be mindful of these developments and use the transition to ensure that broadcasters are responsible stewards of data from the outset. * * * Thank you again for your time today. We’ve come a long way. If we stay committed—and sustain our efforts—believe me, the best is yet to come.