Foreign Ownership of ISPs

I saw a recent announcement that the FCC’s Wireless Competition Bureau had approved the transfer of WideOpenWest to Japan’s Softbank Group. WideOpenWest is the eighth-largest cable company, which operates under the brand name of WOW! The proposed transaction has Softbank buying the DigitalBridge Group, which is WOW!’s majority owner. This deal was announced at the end of 2025, just a few days before DigitalBridge and Crestview Partners took WOW! private.

The blog raises the question about foreign ownership of ISPs because the current administration has put a huge amount of emphasis on America First, which emphasizes, among other things, domestic ownership of corporations doing business in the country. There has been a statutory limit on foreign ownership in U.S. corporations that hold common carrier status or hold broadcast licenses that was created by the Telecommunications Act of 1934. However, the FCC has the ability to waive the 25% limit if it thinks the foreign ownership is in the public interest.

What prompted my question was that the FCC just announced a ban on foreign-made advanced robotic devices. This ban covers all mechanical devices, including autonomous mobile robots, humanoid robots, and quadrupeds that are capable of locomotion, obstacle avoidance, navigation, or movement that operate at a distance from a human operator.

The FCC ban was done by adding foreign-made robots to the Covered List, which is a list the FCC maintains of electronics and devices that are banned from the U.S. At the same time the FCC banned robots, they also banned foreign-made power inverters. At least for now, this ban would even stop the sale of Roombas and other automated vacuums. The FCC knew this ban would be controversial and published a lengthy FAQ about the robot and power-inverter bans. You may recall that the FCC added WiFi modems to the covered list recently.

Assuming that the WOW! transaction will be completed, they won’t be the only foreign-owned ISP in the country. The biggest is T-Mobile, with a majority ownership by the German firm Deutsche Telekom. I wrote a blog recently talking about how Deutsche Telekom is trying to merge with T-Mobile to increase the ability of T-Mobile to expand and grow much larger. It seems likely that if Deutsche Telekom can pull off this merger, the combined company would pursue the acquisition of other large U.S. ISPs. Timotheus Höttges, the CEO of Deutsche Telekom, thinks T-Mobile should be competing on an equal footing with AT&T and Verizon.

Another foreign-owned ISP is Altice USA, which is owned by the French-Israeli billionaire Patrick Drahi, who holds the 74% of shares of stock and 98% of the voting rights of the company. Altice operates under the brand name Optimum and has almost 5 million customers in 21 states. I’ve written several blogs about Altice, which is currently facing a major debt crisis. The company has $21.8 billion in debt, with a payment of $6.2 billion due in 2027. Altice has sued two of its major creditors of the company, Blackrock and Apollo Asset Management, claiming they are blocking the company from refinancing the debt.

Another foreign-owned ISP is Zipley, which is owned by BCE Inc (Bell Canada Enterprises). BCE acquired Zipley for $5 billion in 2025 after receiving a foreign ownership waiver from the FCC.

I don’t have a strong opinion about foreign ownership of ISPs. But I do find it curious that most of the other bans, like robots and WiFi modems, are being done for national security reasons. I can’t think of many industries that we care about more than those that operate and control access to the Internet, and I find it curious that the FCC continues to make exceptions to the 25% foreign ownership cap without much public debate.

Broadband for Low-Income Housing

Connect Humanity the Alt Funding Field Guide that is intended to provide ideas for those who are working to solve the digital divide, like digital navigators. The purpose of the guide is to provide a range of ideas about the immediate future of the digital inclusion field and to provide ideas for practitioners on where to look for funding in an environment where federal grant funding has largely disappeared. I should disclose that I participated in writing one of the chapters.

One of the most interesting chapters was written by Catherine Crago Blanton, who worked with affordable housing in Austin, Texas. She describes the benefits of bringing broadband to low-income housing, both to residents and to landlords.

For those who aren’t familiar with low-income housing, one of the best reasons to read this chapter is to get a peek at the complexity of subsidized housing. The folks who operate low-income housing operate under a wide range of different regulatory rules and federal and state agencies. Most people probably think that most low-income housing is public housing operated by the government, but there is a lot of housing operated by a variety of different kinds of landlords.

This chapter discusses the importance of bringing broadband to low-income housing. We’ve had a slew of different federal and state grants that are putting a big dent in rural broadband access. But those grant programs largely ignored the big remaining population without good broadband, which are folks living in MDUs that don’t have good broadband. A decade ago, much of the entire MDU (multi-dwelling units) market lacked good broadband because neither the ISPs nor landlords wanted to sink the costs into wiring older MDUs. But starting a little over a decade ago, commercial ISPs figured out ways to more affordably wire buildings with fiber or Category 6 cable, and a large majority of MDUs now have decent broadband. In some, ISPs market directly to landlords, and on others, ISPs bring a big data pipe and the landlord builds broadband into the rent.

The one group that has not seen the big burst of ISP activity is low-income MDUs. This market segment thought we had a solution for affordable broadband with the ACP plan that provided a $30 monthly subsidy. That was enough revenue to justify wiring MDUs and building broadband into the rent. But it’s a real challenge to bring broadband to low-income housing without some kind of subsidy. It doesn’t help that HUID rules don’t consider broadband as a necessity and allow it to be covered like other utilities.

This chapter talks about real-life benefits of bringing broadband, and the related digital inclusion efforts of providing computers and computer training into low-income housing. The article discusses examples of how broadband gives tenants a chance to engage with education and workforce training. There is a real-life example of a tenant who had to repeatedly miss work, and the vital income that generated, to take children to doctor visits. There are studies that show that home broadband has a huge positive impact on school children who complete homework in the home environment.

Broadband is increasingly valuable to the older tenants who comprise a lot of subsidized housing. Telehealth brings medical care without the need to arrange for expensive transportation. Home monitoring of health issues results in significantly lower costs and better health for tenants. Overall, broadband can support aging in place, keeping elderly tenants from having to enter the incredibly expensive managed care system.

Landlords have also come to love the benefits of broadband. The folks who manage low-income housing are facing a funding crisis and have to manage properties with fewer resources. When tenants have broadband, they can communicate more effectively with the landlord for issues like repair and maintenance work orders and rent payments. Landlords have come to understand that tenants with broadband are able to respond better and on time to crisis events that result in evictions. Evictions and replacing tenants are a huge cost and effort for landlords.

The Broadband Affordability Crisis

A year ago, John Horrigan wrote an article for the Benton Foundation that predicted that as many as 5.6 million households would drop their home broadband connection due to the impact of the One Big Beautiful Bill. He based that assessment on the changes that the OBBB will make to federal programs that benefit low-income families.

The federal programs that Horrigan thought would hurt broadband subscriptions for low-income households included:

  • The Center on Budget Policy and Priorities estimated that four million people would lose SNAP benefits (food stamps for those not familiar with the acronym) between July 2025 and March 2026. The actual number turned out to be 4.3 million.
  • Horrigan predicted that millions would lose a federal healthcare subsidy from ACA, or the various state programs that are part of the ACA. So far, enrollees in ACA programs have dropped from 22.1 million in 2025 to 19.2 million – a drop of 2.9 million.
  • There was a prediction that insurance premiums would increase for everybody due to the drop in ACA enrollments. The 19.2 million people that stayed in the ACA health system saw an average monthly premiums increase of $122 per month. The average health insurance premium for the country as a whole increased by $178, while the full out-of-pocket costs increased even more due to increases in deductibles and copayments.

Now that we are halfway through 2026, there are also cost increases hitting households that Horrigan didn’t anticipate. The nationwide average cost of a gallon of gas was $2.81 at the end of 2025 and is now over $4.00. There are a number of estimates showing that the monthly cost of groceries for a family has increased by $30 to $45. The real big upcoming kicker is that health insurance rates, both for ACA participants and everybody else, are expected to be significantly higher in 2027 than they were this year. It seems there will also be a big cut in federal support for Medicare prescriptions.

What does all of this have to do with broadband subscriptions? There are currently over 26 million homes in the U.S. in the lowest quintile that have household incomes below $33,000 per year. For these households, increases in expenses for any of the above reasons mean they have to cut expenditures elsewhere. Somebody who loses SNAP or who pays more for gas or insurance has to make up by cutting elsewhere.

Horrigan’s prediction of 5.6 million homes that will lose broadband started with a universe of homes that already have broadband. When a household has an income under $33,000 per year and is buying broadband, they think it’s important.

Horrigan’s prediction may not fully come to pass because there is a middle-ground to abandoning broadband. Earlier this year, I worked with two ISPs that serve areas that are primarily low-income. Both ISPs told me that they are experiencing a big increase in a new kind of churn – customers that temporarily drop broadband because they can’t afford it, but who then resubscribe a few months later when they’ve scraped together enough money to buy broadband again.

The other thing that a cost squeeze does is to drive homes to look for broadband alternatives. I think that people in this category are the ones dropping cable companies and going to FWA broadband and other less expensive options. Unfortunately for families in this position, lower monthly rates typically mean slower speeds and less reliable broadband performance. A lot of low-income families have stayed with DSL for years because it was the least costly option in cities. However, in recent years, the big telcos have increased the cost of DSL to be the same price as fiber to get people off the copper network – but these higher prices are often too high for too many homes with low incomes.

There is a nationwide consensus that we are all better off when most families have broadband and can partake in the many benefits broadband brings to a household. If we don’t somehow change the current path we are on, I think Horrigan’s estimate of 5.6 million households that drop broadband will be far too low.

Is Satellite Cellular a Competitive Threat?

It seems that one of the hottest topics in the industry is Direct-to-Device (D2D) cellular service that can bypass terrestrial cell towers and connect cellphones directly to satellites. I’ve already seen claims being made that D2D technology will solve the rural cellular issue since everybody will have cell coverage. Today’s blog looks at some of the reality behind these claims.

There are a lot of players suddenly in this space. Not surprisingly, SpaceX / Starlink already has hundreds of satellites in orbit and plans for 15,000. AST SpaceMobile is partnering with AT&T and Verizon. Globalstar originally pioneered Apple’s SOS emergency test service and was recently acquired by Amazon, which recently applied for licenses for 5,105 satellites for D2D. Lynk Global has partnered with more than 45 cellular companies worldwide. EchoStar and Viasat have also announced plans to enter the market.

Basic Technology. The satellites will circle at heights between 200 and 550 miles above Earth and will be traveling at 17,000 miles per hour. At least for now, nobody has figured out how to hand off calls between a satellite and terrestrial cellular network. There are some limitations for D2D that are largely defined by physics.

  • Can’t Work Indoors. Your cellphone transmits at under 0.5 watts of power – the maximum power level set by the FCC. Cell signals lose from 50% to 90% of the signal strength when passing through a wall. While that usually is enough power to reach a cell tower located within 2-3 miles of your house, it’s not nearly enough power to reach a satellite.
  • Receiver Limits. Starlink has been achieving data speeds up to 300 Mbps download for home broadband. Expect speeds to be far less for a D2D due to the much smaller antenna surface area in a cellphone compared to a Starlink receiver dish. In general, the larger the size of the antenna, the more bits that can be gathered. Receivers for Starlink range from 11.75 X 15.7 inches for the Starlink Mini to 22.7 X 20.1 for the higher-performance receiver. The typical cellphone has a receiver of around 6.5 X 7 millimeters, meaning that even the Starlink Mini has 2,600 times more surface area for receiving a signal from the satellite than a typical cellphone.
  • Area Speed Coverage Limits. The planned coverage area for the next generation of D2D Starlink satellite is around 45 square miles. The newer satellites will be using massive adjustable antenna beams along with 40 to 65 MHz channels. This means that a single data stream from a satellite will have an overall data capacity between 40 and 150 Mbps download, depending on the spectrum being used. Starlink is planning on 16 simultaneous full-power beams from the next generation D2D satellites. The challenge is that this spectrum must be shared among all users inside a beam footprint. That means a maximum of 2.4 Gbps of bandwidth to share in a 45 square mile area – a minuscule amount of bandwidth compared to what can be delivered by cell towers serving the same area. This is compounded by a limitation on the simultaneous number of users that can be connected at the same time.

What does this all mean from a practical perspective? The D2D satellites will be able to connect to a large number of people who are outdoors, within the overall limitation of the number of simultaneous people that can be connected at one time. We won’t know until somebody shows us what this will mean for connecting a lot of people in a small area, like at a rural rock concert.

Early trials have shown that D2D has some capacity for making a connection to a moving vehicle. This is never going to be as guaranteed as when standing outdoors, and it will likely depend on the type of vehicle and the speed being driven.

D2D is not going to work directly indoors. It’s possible that somebody will develop an outdoor receiver that can bounce a signal to go in through a window. It would make sense to do this in conjunction with something like a Starlink receiver.

It’s hard to envision D2D as a standalone cellular service other than possibly for the rare person who spends most of their time outdoors.

To be fair to the industry, at least for now, the big D2D providers aren’t touting it as a standalone cellular service. They are talking about this being an add-on to normal cellular plans. This doesn’t make much sense in urban areas where normal cell coverage is already ubiquitous and strong. But this could be an interesting additive to a cell plan for folks who either live or travel a lot in rural areas. It’s a cool technology, but a lot of the excitement could be empty hype, which seems to be the norm for most new technology products.

There is only one statistic that is really going to matter for D2D, which is the number of people willing to pony up a monthly fee to use it. The one thing going for the new technology is that there is a lot more serious demand for any kind of cellular coverage in the third world than there is in the rural U.S. Maybe that’s the real market they are shooting for.

Data Sovereignty

Lately I’ve been hearing a lot about data sovereignty. Indian Tribes want data sovereignty. Countries around the world and state and local governments in the U.S. are talking about data sovereignty. Corporations are talking about data sovereignty, even though they use different terms to describe it.

The concept behind data sovereignty is simple. It means keeping data internally and not sharing it with the outside world, which today means keeping it away from the AI companies that are training their models on every bit of data they can put their hands on.

Consider corporations. I’m positive that a huge amount of what happens inside corporations ends up in the AI cloud. It seems inevitable that employees in every company are relying on AI to help them compose emails, analyze budgets, and write memos. Employees use AI notetakers to capture every word said in meetings. Even if a company has a policy of not doing any of this kind of work using public AI, you know employees are doing so privately to make their jobs easier.

Entire countries are talking about data sovereignty, particularly in Europe. There is a quickly growing top-down movement on the continent to ban and replace large U.S. software tech vendors. We’ve recently seen the governments in France, Spain, Germany, the UK, and Switzerland begin the process to get rid of Palantir, a U.S. surveillance software vendor, and replace it with European software. Each country doesn’t only want to protect data from U.S. tech companies, but also from each other. There is a lot of discussion in Europe about replacing many other U.S. tech firms.

The data sovereignty effort in the U.S. is much more of a bottoms-up movement. Many city government CIOs are demanding to have local control of citizens’ data. This partly comes from a rash of cities and local governments that have been hacked by ransomware attacks in the last few years. But this is also in response to citizens demanding that local governments protect their data.

Of course, there is opposition to the data sovereignty idea. For example, there is a tech coalition that has been pushing the Administration to create “freedom cities” that are exempt from federal regulations and where tech companies would have free rein. An effort to do this in Toronto failed a few years ago when citizens rebelled against the idea of creating a high-tech surveillance community.

Indian Tribes might be ahead of everybody else on this issue, and there is a lot of discussion of keeping all tribal data sovereign from the rest of the world. Tribes have no interest in the federal government getting their local data and are even less interested in seeing tech companies profiting by selling the data to anybody willing to buy it.

The idea of data sovereignty runs completely counter to the idea of building huge data centers that amass all of the knowledge of mankind. Sovereignty is really about privacy and logically leads to an AI environment comprised of local AI servers for corporations, local governments, and Tribes that are not connected to the cloud or to the big AI data centers.

Interestingly, the data sovereignty movement is not anti-AI. It instead comes from those who are against having their data gobbled up by a handful of big tech companies that want to monetize their data with no care for what that means to the people, companies, and governments that generate the most useful data. It’s likely that most entities that shield themselves from big AI will use AI internally inside their own network. I think a whole lot of people have already seen the massive invasion of privacy that is an integral part of AI and are already looking for ways to opt out.

Local Governments and ISPs

Today’s blog is a tale of two counties that were successful in attracting broadband grant investments and now want to pivot to making sure that residents are able to take advantage of the newly constructed fiber networks. Both counties are encountering an unexpected roadblock.

Both counties that did things the right way. Both counties put a lot of effort into attracting better broadband. They funded major feasibility studies. These studies included activities like sponsoring well-attended public meetings, doing surveys to understand the interest in broadband, and exploring all of the local issues that ISPs want to know about when they are considering building new networks. Most importantly, both counties used some of their own federal ARPA funding to make broadband grants to ISPs, which served as seed funding for ISPs looking for state and federal grant money to build fiber networks.

And the efforts by the counties paid off. Both counties attracted ISPs that were successful in finding the grant funding needed to build fiber in rural areas. The fiber construction is mostly completed, and the networks are operational. One county partnered with two large ISPs, a cable company and a telco, and the second county ended up with fiber being built by five, mostly local, ISPs.

Interestingly, both counties realize that infrastructure is only the first step and the work in their counties is not going to be complete until residents who want broadband are connected. I think they need to be applauded for this realization, because I know many counties have assumed that once the networks are built, they can move their focus to other issues.

The reason behind this blog is that both counties say they have reached the same roadblock, in that the ISPs that were funded to build fiber won’t tell them how many customers are subscribing to the new fiber networks. This caught both counties by surprise for several reasons. First, they thought they had forged a relationship with the ISPs during the process of awarding local grants and through the general process of working together to attract more grants. More surprising to them is that each county is willing to put effort and funding into getting more residents online by supporting a variety of digital inclusion efforts like getting more computers into homes and teaching people how to use online resources.

The folks at the counties assumed that digital inclusion efforts are a win:win:win for the residents, the ISPs, and the counties. The counties understand that getting people online means better opportunities for education and higher pay. They assumed the ISPs would be interested in getting more customers. But they are getting nowhere with the ISPs. ISPs ranging from small local providers to large national cable companies gave them the same response.

I know ISPs well, and I’m not surprised by the response. In a competitive world, ISPs don’t want to give subscriber information to a local government which they assume will become public – and they are probably right about that.

I’ve spent a few days scratching my head about this and wondering how we break this particular log jam, which I assume is being repeated all across the country. There are local governments and digital inclusion advocates of all types who are willing to put in the effort to get people connected to broadband. ISPs clearly benefit by getting more customers. But many ISPs are leery of looking outside the company for help, and the idea of trusting outsiders to bring new customers is a foreign idea.

I know there are ISPs that have embraced digital inclusion efforts. However, there are surprisingly few stories about this in circulation. Maybe that silence is on purpose, and these efforts are being tackled locally with little fanfare. I’d love to hear from anybody who has figured out how to break the logjam being experienced by these counties.

Technology-neutral Policy

Arielle Roth, the Administrator of NTIA, made a speech to the American Enterprise Institute that was titled There Is No Such Thing as “Future-Proof” Technology. Her speech was in reference to the upcoming 6G technology.

She made a really interesting observation about 6G that I hadn’t seen before: That lesson is especially timely as we stand at the frontier of AI-native 6G . . . Perhaps the defining feature of AI-native 6G is not simply that it will make communications networks more intelligent. It is that it will make the future of communications even less predictable. Previous generations of wireless networks were designed primarily to move information faster. The next generation promises networks that increasingly optimize, adapt, and derive insight from the information they carry. As intelligence becomes embedded within the network itself, we should become more humble about our ability to know which technologies, architectures, or applications will ultimately prevail.

Obviously, NTIA is on the 6G bandwagon, much like the first Trump administration was on the 5G bandwagon. To some degree, it’s the government’s role to promote new technologies, and this is done for many other technologies outside of telecom.

I have my doubts that carriers are going to build networks that are less predictable, because that sounds rife with risk. I have serious doubts that the big carriers are going to hand off the responsibility to operate 6G networks completely to AI. The industry learned a lesson last year with the big Verizon outage that lasted far longer than it should have when technicians couldn’t figure out how to fix problems in AI-generated software. Everything I’ve been reading is that the big nationwide networks are bringing more human intervention back into the operation and troubleshooting of networks.

The speech then went on to talk about technology-neutrality. Roth said that the government should never favor any technology, since doing so changes the direction of industry by crowding out innovation.

To some degree, it’s hard to argue with that statement since there are many examples in the past where governments stuck with older technologies and ignored innovations. But it’s hard to say that the current Administration is technology-neutral when it favors traditional power technologies over wind and solar.

Roth said that technology-neutrality matters the most in rural areas, where low-Earth orbit satellite constellations have reduced the cost of reaching people. She said satellite has made obsolete the idea that we need to build wireline networks to reach homes.

I don’t think she is going to find very many local governments and people in rural areas who share her vision that satellite technology is superior to fiber networks. I’ve worked with several hundred counties who understand that a fiber network build today puts their county in a good competitive position for the rest of this century. I don’t know one rural county who regrets the construction of fiber networks.

I find it interesting that she brought up rural broadband in the context of touting 6G, because 6G is not coming to most of rural America, just like 5G didn’t. If anything, the transition from using lower frequencies for 4G to higher ones for 5G and 6G means that rural cellular coverage will continue to shrink over time.

Just as an aside, as she has done in the past, Roth referred to the reduction of BEAD infrastructure grants as a savings to the taxpayer – again hinting that NTIA is going to be reluctant to release much, if any, of the BEAD nondeployment funds.

After the AI Crash

Everything I read about the AI industry leads me to think there will be an AI crash. Consider the following:

  • Unsustainable Capital Expenses. It’s hard to imagine there can ever be enough revenue to pay for the huge capital investments in data centers and electronics. Several analysts have estimated that it will take $2 trillion a year in revenue to pay for the infrastructure that has already been built, and there are no believable forecasts for generating even half that much revenue. The capital needs of the industry are relentless since expensive AI data center electronics have to be replaced within five years, or less.
  • Circular Revenues. A small handful of tech firms, chip manufacturers, and AI companies are propping each other up by investing and buying from each other. If one stumbles, they might all fall.
  • Huge Debt. Much of the industry is being funded through debt, which has to eventually be repaid, instead of through equity.
  • Public Pushback. Local governments and people are increasingly pushing back hard against the creation of new data centers. Most new technologies have been welcomed by the public with open arms.
  • Increasing Corporate Skepticism. The news is full of stories of corporations that are throttling the employee use of AI since the costs to use the software are a lot higher than expected. There are many companies having second thoughts about replacing people with AI. The AI industry needs complete corporate buy-in to have any chance of succeeding, and large companies are generally still on the sidelines.
  • Diseconomies of Scale. Every new technology I can think of thrived, in part, due to economies of scale, where the larger the industry grew, the more efficient it got. AI is going in the opposite direction, where every new AI model consumes more resources than its predecessors. This may turn out to be the fatal flaw – the bigger the industry gets, the more its operating costs increase.
  • Institutional Warnings. Moody’s recently warned that high AI infrastructure spending threatens the credit of AI companies and their large tech partners. I read recently that the number one question being fielded by investment advisors is people asking how to divest from AI.

I don’t have a crystal ball to foresee the nature of the crash. It could be a total crash like the 2000 tech crash, where four out of five tech startups disappeared practically overnight. I lived in the DC area at the time, and I will never forget the rows of abandoned CLEC headquarters buildings in Northern Virginia. A crash could be milder, where a few firms disappear, with the outlooks for the survivors greatly diminished, and industry expectations are reset to something more realistic.

The reason I wrote the blog is to speculate about what happens after an AI crash. I foresee some of the following consequences of an AI crash.

  • An article in the Economist said a total crash would wipe out $20 trillion in U.S. wealth. That means wiping out the wealth of the investors in the new technology, along with a huge hit on the stock market.
  • Data center construction would stop dead, and unfinished projects would collapse. Communities that contributed to the costs of bringing data centers will end up eating those investments.
  • There will be stranded investments by electric utilities and water companies that built new infrastructure to support data centers. They won’t eat these losses, though, which will all be passed on to ratepayers in the form of higher electric and water rates.
  • A lot of vendors will be in big trouble. Companies that pivoted to supporting data center electronics, like Micron, might fold. But a lot of other vendors also would take a big hit. For example, Corning announced investments in three new fiber factories just to support data centers.
  • There have been some huge investments by carriers in middle-mile fiber to support data centers. The companies that made these investments won’t see the expected revenues.

The most interesting thing about a major crash is that it can do as much long-term good as it does short-term harm. I want to again use the analogy from the tech crash. I know of at least a half dozen CLECs that had business plans to capture 30% of the voice and data market in Atlanta. The crash cleaned them all out of the market, but without the crash they would have all failed more slowly. The tech crash brought a sense of reality to the telecom market, which still experienced phenomenal long-term growth after the original tech companies had died.

I don’t think there is any chance of AI failing as a technology. But that doesn’t mean the early developers are the ones who will see the ultimate success. Most, and maybe all of today’s players might be gone. A crash will bring financial constraints, which would mean that AI companies will have to figure out efficiency and economies of scale. If AI is ever going to be a viable technology, it has to control costs and be able to pay for itself. It’s hard to foresee today’s companies somehow reaching that point without some kind of market reset.

A New Strategy for the FCC

The FCC issued a Strategic Plan for 2026 to 2030 that describes a significant change in the focus of the agency. The FCC has periodically issued strategic plans since it was directed to do so by the Government Performance and Results Act in 1997. The strategic plan doesn’t change any FCC rules. It’s more of an outline of how the agency views its function and defines its priorities going forward. The plan has one formal use, which is to define the yardstick by which the GAO will measure whether the FCC is doing what it said it would do.

https://www.fcc.gov/document/strategic-plan-2026-2030

The new plan clearly represents Chairman Carr’s priorities, which he has been espousing since he took the Chairman position. Chairman Carr characterizes his policy initiatives as a Build America Agenda that includes:

  • Focusing on spectrum as a way to create jobs, increase competition, and drive down prices.
  • Focusing on the space economy to make sure that America is the world leader in this new industry.
  • Reducing red tape and making it easier for carriers to deploy infrastructure investments.
  • Focusing on national security to make sure that our networks are resilient and secure.
  • Protecting free speech.
  • Being a good steward of taxpayer money.

This is a huge departure from the Strategic Plan issued by former FCC Chairperson Jessica Rosenworcel. That past plan focused on bringing high-speed Internet to everybody, with a focus on affordability, digital equity and adoption, and consumer protection. The new plan almost entirely ignores the goals of the last FCC. One of the few areas of agreement between the two plans is the effort to eliminate robocalls and spam.

We’ve already seen all of the priorities of this plan in action.

  • The FCC is laser-focused on expanding mid-band spectrum to support 5G/6G. This was partly driven by direction from Congress in the One Big Beautiful Bill that directed the FCC to find 800 MHz of new spectrum for auction, but it was already a priority of Chairman Carr on day one. The FCC recently completed an auction for AWS-3 spectrum and is working towards an auction for Upper C-Band spectrum.
  • It’s clear that the FCC is focused on space. The FCC recently freed up new spectrum for satellite broadband and is considering even more spectrum. The FCC recently relaxed regulations related to approving new satellite constellations and ventures.
  • Much of the FCC’s actions this year have been aimed at making it easier for carriers to deploy infrastructure. The FCC’s tactics have mostly been aimed at overriding local and state authority and regulation of issues like wireless site placement, pole attachments, and permitting. The FCC also eliminated regulations and cleared barriers to removing copper networks. One of the first actions of this FCC was the Delete, Delete, Delete effort to eliminate outdated regulations.
  • In the area of national security, the FCC has tightened up rules about carriers using modems and other foreign-made devices and is discussing lowering the dependence on foreign technology in a wide range of the industry.
  • The free speech issue has puzzled me from day one, since Chairman Carr seems to be focused on punishing TV networks and others who don’t agree with the Administration. This seems like the opposite of free speech, but so far it’s mostly been rhetoric with little actual FCC action.
  • In the areas of being a good steward of taxpayer money, the FCC has focused on perceived fraud in Universal Service programs and is in the middle of a major investigation of all parts of the USF.

The plan is worth reading because it describes how Chairman Carr thinks about every aspect of what the FCC is doing. The plan puts the actions of the agency into perspective and defines the overall regulatory philosophy. This is an FCC that clearly values the big companies it regulates over the public and consumers. It’s not the first FCC to do so, and the document spells it out for anybody who reads it.

SpaceX Wants to Handle the World’s Broadband

In a recent request filed with the FCC, Starlink asked to increase the size of its broadband satellite fleet to 100,000. This coincided with the FCC’s announcement that it wants to ease the process for approving new satellites, so I have to think the idea will sail through. There was one extraordinary statement in this filing that is the topic of today’s blog. SpaceX said it has the goal of handling the majority of the world’s Internet traffic. Elon Musk has always been one for huge hyperbole, so statements like this are not surprising. I want to examine a few reasons why that goal is ridiculous.

No carrier handles a large percentage of the world’s internet traffic today. According to statistics shown on Cloudflare Radar, the company with the biggest percentage of worldwide traffic today is Amazon, with 3.5% of all traffic. The companies that carry more than 2% of worldwide traffic include Cloudflare (2.2%), Comcast (2.5%), Amazon (2.4%), Microsoft (2.2%), Reliance/Jio (2.1%), and AT&T (2.1%). SpaceX is on the worldwide list at number 18 today, already carrying 0.7% of the world’s Internet traffic. SpaceX would have to grow by five times to catch Amazon. But that would still be a long way from carrying a majority of the world’s traffic, and it’s hard to imagine any carrier growing to carry even 10% of the world’s traffic.

Starlink has built an impressive ISP business by gaining 12 million broadband customers. Worldwide, there are over 6 billion Internet users, about 74% of the world’s population. If you assume an average of 2.5 people per household, Starlink today represents 0.5% of all residential Internet users. But residential broadband, as large as it is, is still only a fraction of the world’s Internet traffic. Again, according to Cloudflare, bots are currently making 56.8% of all web connections.

Starlink largely has a near-monopoly for satellite broadband today, but that’s going to change over the next five years. There are major plans by others to compete with Starlink, including Amazon, China’s Guowang, Shanghai’s Quinfan, Blue Origin’s TeraWave, and Eutelsat’s OneWeb. Competition will dilute Starlink’s share of the satellite market.

Love him or hate him, Elon Musk has aligned himself with politics, which is unusual for somebody operating a giant retail business. There are already many countries that ban Starlink, and this list could grow. Even in countries where Starlink is available, there are a lot of people who would never buy service from an Elon Musk-backed company.

I think the biggest reason is that there is a lot more risk associated with trusting your broadband to satellites. A massive 1,000-year solar flare could knock a lot of satellites out of commission. There is a real threat of a spreading cloud of space debris killing a lot of satellites, as predicted by the Kessler syndrome. There is a chance that satellites could be knocked out deliberately during a war or by a rogue nation.

Fiber networks face no similar worldwide threats. Large users of broadband are never going to trust their traffic to satellite. Don’t expect to see data centers, universities, governments, or large corporations favor satellite connections over fiber.

I have to admit that I like Elon Musk’s brashness, but his claims are always overboard. I think he has a better chance of establishing his colony on Mars than of capturing the majority of the world’s Internet traffic.