At RightsCon in Taipei, activists reckon with a US retreat from promoting digital rights

Stay Ahead, Stay ONMINE

At RightsCon in Taipei, activists reckon with a US retreat from promoting digital rights

Last week, I joined over 3,200 digital rights activists, tech policymakers, and researchers and a smattering of tech company representatives in Taipei at RightsCon, the world’s largest digital rights conference. Human rights conferences can be sobering, to say the least. They highlight the David vs. Goliath situation of small civil society organizations fighting to center human rights in decisions about technology, sometimes challenging the priorities of much more powerful governments and technology companies. But this year’s RightsCon, the 13th since the event began as the Silicon Valley Human Rights Conference in 2011, felt especially urgent. This was primarily due to the shocking, rapid gutting of the US federal government by the Elon Musk–led DOGE initiative, and the reverberations this would have around the world. At RightsCon, the cuts to USAID were top of mind: the development agency has long been one of the world’s biggest funders of digital rights work, from ensuring that the internet stays on during elections and crises around the world to supporting digital security hotlines for human-rights defenders and journalists targeted by surveillance and hacking. Now, the agency is facing over 90% cuts to its budget under the Trump administration. The withdrawal of funding is existential for the international digital rights community—and follows other trends that are concerning for those who support a free and safe Internet. “We are unfortunately witnessing the erosion … of multistakeholderism, with restrictions on civil society participation, democratic backsliding worldwide, and companies divesting from policies and practices that uphold human rights,” Nikki Gladstone, RightsCon’s director, said in her opening speech. Cindy Cohn, director of the Electronic Frontier Foundation, which advocates for digital civil liberties, was more blunt: “The scale and speed of the attacks on people’s rights is unprecedented. It’s breathtaking,” she told me. But it’s not just funding cuts that will curtail digital rights globally. As various speakers highlighted throughout the conference, the United States government has gone from taking the leading role in supporting an open and safe internet to demonstrating how to dismantle it. Here’s what speakers are seeing: The Trump administration’s policies are being weaponized in other countries On Tuesday, February 25, just before RightsCon began, Serbian law enforcement raided the offices of four local civil society organizations focused on government accountability, citing Musk and Trump’s (unproven) accusations of fraud at USAID. “The (Serbian) Special Anti-Corruption Department … contacted the US Justice Department for information concerning USAID over the abuse of funds, possible money laundering, and the improper spending of American taxpayers’ funds in Serbia,” Nenad Stefanovic, a state prosecutor, explained on a TV broadcast announcing the move. “Since Trump’s second administration, we cannot count on them [the platforms] to do even the bare minimum anymore.” -Yasmin Curzi For RightsCon attendees, it was a clear—and familiar—example of how oppressive regimes find or invent reasons to go after critics. Only now, by using the Trump administration’s justifications for revoking USAID’s funding, they hope to gain an extra veneer of credibility. Ashnah Kalemera, a program manager for CIPESA, a Ugandan nonprofit that runs technology for civic participation initiatives across Africa, says Trump and Musk’s attacks on USAID are providing false narratives that “justify arrests, intimidations, and continued clampdowns on civil society organizations—organizations that obviously no longer have the resources to do their work anyway.” Yasmin Curzi, a professor at FGV Law School in Rio de Janeiro and an expert on digital law, says that American politics are also being weaponized in Brazil’s domestic affairs. There, she told me, right wing figures have been “lifting signs at protests like ‘Trump save us!’ and ‘Protect our First Amendment rights,’ which they don’t have.” Instead, Brazil’s Internet Bill of Rights seeks to balance protections on user privacy and speech with criminal liabilities for certain types of harmful content, including disinformation and hate speech. Despite the differing legal frameworks, in late February the Trump Media & Technology Group, which operates Truth Social, and the video platform Rumble tried to enforce US-style speech protections in Brazil. They sued Brazilian Supreme Court Justice Alexandre de Moraes for banning a Brazilian digital influencer who had fled to the United States to avoid arrest in connection with allegations that he has spread disinformation and hate. Truth Social and Rumble allege that Moraes has violated the United States’ free speech laws. (A US judge has since ruled that because the Brazilian court had yet to officially serve Truth Social and Rumble as required under international treaty, the platforms’ lawsuit was premature and the companies do not have to comply with the order; the judge did not comment on the merits of the argument, though the companies have claimed victory.) Platforms are becoming less willing to engage with local communities In addition to how Trump and Musk might inspire other countries to act, speakers also expressed concern that their trolling and use of dehumanizing language and imagery will inspire more online hate (and attacks), just at a time when platforms are rolling back human content moderation. Experts warn that automated content moderation systems trained on English-language data sets are unable to detect much of this hateful language. India, for example, has a history of platforms’ recognizing the necessity of using local-language moderators and also failing to do so, leading to real-world violence. Yet now the attitude of some internet users there has become “If the president of the United States can do it, why can’t I?” says Sadaf Wani, a communications manager for IT for Change, an Indian nonprofit research and advocacy organization, who organized a RightsCon panel on hate speech and AI. As her panel noted, these online attacks are accompanied by an increase in automated and even fully AI-based content moderation, largely trained on North American data sets, that are known to be less effective at identifying problematic speech in languages other than English. Even the latest large language models have difficulties identifying local slang, cultural context, and the use of non-English characters. “AI is not as smart as it looks, so you can use very obvious [and] very basic tricks to evade scrutiny. So I think that’s what’s also amplifying hate speech further,” Wani explains. Others, including Curzi from Brazil and Kalemera from Uganda, described similar trends playing out in their countries—and they say changes in platform policy and a lack of local staff make content moderation even harder. Platforms used to have humans in the loop whom users could reach out to for help, Curzi said. She pointed to community-driven moderation efforts on Twitter, which she considered to be a relative success at curbing hate speech until Elon Musk bought the site and fired some 4,400 contract workers—including the entire team that worked with community partners in Brazil. Curzi and Kalemera both say that things have gotten worse since. Last year, Trump threatened Meta CEO Mark Zuckerberg with “spend[ing] the rest of his life in prison” if Meta attempted to interfere with—i.e. fact-check claims about—the 2024 election. This January Meta announced that it was replacing its fact-checking program with X-style community notes, a move widely seen as capitulation to pressure from the new administration. Shortly after Trump’s second inauguration, social platforms skipped a hearing on hate speech and disinformation held by the Brazilian attorney general. While this may have been expected of Musk’s X, it represented a big shift for Meta, Curzi told me. “Since Trump’s second administration, we cannot count on them [the platforms] to do even the bare minimum anymore,” she adds. Meta and X did not respond to requests for comment. The US’s retreat is creating a moral vacuum Then there’s simply the fact that the United States can no longer be counted on to support digital rights defenders or journalists under attack. That creates a vacuum, and it’s not clear who else is willing—or able—to step into it, participants said. The US used to be the “main support for journalists in repressive regimes,” both financially and morally, one journalism trainer said during a last-minute session added to the schedule to address the funding crisis. The fact that there is now no one to turn to, she added, makes the current situation “not comparable to the past.” But that’s not to say that everything was doom and gloom. “You could feel the solidarity and community,” says the EFF’s Cohn. “And having [the conference] in Taiwan, which lives in the shadow of a very powerful, often hostile government, seemed especially fitting.” Indeed, if there was one theme that was repeated throughout the event, it was a shared desire to rethink and challenge who holds power. Multiple sessions, for example, focused on strategies to counter both unresponsive Big Tech platforms and repressive governments. Meanwhile, during the session on AI and hate-speech moderation, participants concluded that one way of creating a safer internet would be for local organizations to build localized language models that are context- and language-specific. At the very least, said Curzi, we could move to other, smaller platforms that match our values, because at this point, “the big platforms can do anything they want.” Do you have additional information on how Doge is affecting digital rights globally? Please use a non-work device and get in touch at [email protected] or with the reporter on Signal: eileenguo.15.

Human rights conferences can be sobering, to say the least. They highlight the David vs. Goliath situation of small civil society organizations fighting to center human rights in decisions about technology, sometimes challenging the priorities of much more powerful governments and technology companies.

But this year’s RightsCon, the 13th since the event began as the Silicon Valley Human Rights Conference in 2011, felt especially urgent. This was primarily due to the shocking, rapid gutting of the US federal government by the Elon Musk–led DOGE initiative, and the reverberations this would have around the world.

At RightsCon, the cuts to USAID were top of mind: the development agency has long been one of the world’s biggest funders of digital rights work, from ensuring that the internet stays on during elections and crises around the world to supporting digital security hotlines for human-rights defenders and journalists targeted by surveillance and hacking. Now, the agency is facing over 90% cuts to its budget under the Trump administration.

The withdrawal of funding is existential for the international digital rights community—and follows other trends that are concerning for those who support a free and safe Internet. “We are unfortunately witnessing the erosion … of multistakeholderism, with restrictions on civil society participation, democratic backsliding worldwide, and companies divesting from policies and practices that uphold human rights,” Nikki Gladstone, RightsCon’s director, said in her opening speech.

Cindy Cohn, director of the Electronic Frontier Foundation, which advocates for digital civil liberties, was more blunt: “The scale and speed of the attacks on people’s rights is unprecedented. It’s breathtaking,” she told me.

But it’s not just funding cuts that will curtail digital rights globally. As various speakers highlighted throughout the conference, the United States government has gone from taking the leading role in supporting an open and safe internet to demonstrating how to dismantle it. Here’s what speakers are seeing:

The Trump administration’s policies are being weaponized in other countries

On Tuesday, February 25, just before RightsCon began, Serbian law enforcement raided the offices of four local civil society organizations focused on government accountability, citing Musk and Trump’s (unproven) accusations of fraud at USAID.

“The (Serbian) Special Anti-Corruption Department … contacted the US Justice Department for information concerning USAID over the abuse of funds, possible money laundering, and the improper spending of American taxpayers’ funds in Serbia,” Nenad Stefanovic, a state prosecutor, explained on a TV broadcast announcing the move.

“Since Trump’s second administration, we cannot count on them [the platforms] to do even the bare minimum anymore.” -Yasmin Curzi

For RightsCon attendees, it was a clear—and familiar—example of how oppressive regimes find or invent reasons to go after critics. Only now, by using the Trump administration’s justifications for revoking USAID’s funding, they hope to gain an extra veneer of credibility.

Ashnah Kalemera, a program manager for CIPESA, a Ugandan nonprofit that runs technology for civic participation initiatives across Africa, says Trump and Musk’s attacks on USAID are providing false narratives that “justify arrests, intimidations, and continued clampdowns on civil society organizations—organizations that obviously no longer have the resources to do their work anyway.”

Yasmin Curzi, a professor at FGV Law School in Rio de Janeiro and an expert on digital law, says that American politics are also being weaponized in Brazil’s domestic affairs. There, she told me, right wing figures have been “lifting signs at protests like ‘Trump save us!’ and ‘Protect our First Amendment rights,’ which they don’t have.” Instead, Brazil’s Internet Bill of Rights seeks to balance protections on user privacy and speech with criminal liabilities for certain types of harmful content, including disinformation and hate speech.

Despite the differing legal frameworks, in late February the Trump Media & Technology Group, which operates Truth Social, and the video platform Rumble tried to enforce US-style speech protections in Brazil. They sued Brazilian Supreme Court Justice Alexandre de Moraes for banning a Brazilian digital influencer who had fled to the United States to avoid arrest in connection with allegations that he has spread disinformation and hate. Truth Social and Rumble allege that Moraes has violated the United States’ free speech laws.

(A US judge has since ruled that because the Brazilian court had yet to officially serve Truth Social and Rumble as required under international treaty, the platforms’ lawsuit was premature and the companies do not have to comply with the order; the judge did not comment on the merits of the argument, though the companies have claimed victory.)

Platforms are becoming less willing to engage with local communities

In addition to how Trump and Musk might inspire other countries to act, speakers also expressed concern that their trolling and use of dehumanizing language and imagery will inspire more online hate (and attacks), just at a time when platforms are rolling back human content moderation. Experts warn that automated content moderation systems trained on English-language data sets are unable to detect much of this hateful language.

India, for example, has a history of platforms’ recognizing the necessity of using local-language moderators and also failing to do so, leading to real-world violence. Yet now the attitude of some internet users there has become “If the president of the United States can do it, why can’t I?” says Sadaf Wani, a communications manager for IT for Change, an Indian nonprofit research and advocacy organization, who organized a RightsCon panel on hate speech and AI.

As her panel noted, these online attacks are accompanied by an increase in automated and even fully AI-based content moderation, largely trained on North American data sets, that are known to be less effective at identifying problematic speech in languages other than English. Even the latest large language models have difficulties identifying local slang, cultural context, and the use of non-English characters. “AI is not as smart as it looks, so you can use very obvious [and] very basic tricks to evade scrutiny. So I think that’s what’s also amplifying hate speech further,” Wani explains.

Others, including Curzi from Brazil and Kalemera from Uganda, described similar trends playing out in their countries—and they say changes in platform policy and a lack of local staff make content moderation even harder. Platforms used to have humans in the loop whom users could reach out to for help, Curzi said. She pointed to community-driven moderation efforts on Twitter, which she considered to be a relative success at curbing hate speech until Elon Musk bought the site and fired some 4,400 contract workers—including the entire team that worked with community partners in Brazil.

Curzi and Kalemera both say that things have gotten worse since. Last year, Trump threatened Meta CEO Mark Zuckerberg with “spend[ing] the rest of his life in prison” if Meta attempted to interfere with—i.e. fact-check claims about—the 2024 election. This January Meta announced that it was replacing its fact-checking program with X-style community notes, a move widely seen as capitulation to pressure from the new administration.
Shortly after Trump’s second inauguration, social platforms skipped a hearing on hate speech and disinformation held by the Brazilian attorney general. While this may have been expected of Musk’s X, it represented a big shift for Meta, Curzi told me. “Since Trump’s second administration, we cannot count on them [the platforms] to do even the bare minimum anymore,” she adds. Meta and X did not respond to requests for comment.

The US’s retreat is creating a moral vacuum

Then there’s simply the fact that the United States can no longer be counted on to support digital rights defenders or journalists under attack. That creates a vacuum, and it’s not clear who else is willing—or able—to step into it, participants said.

The US used to be the “main support for journalists in repressive regimes,” both financially and morally, one journalism trainer said during a last-minute session added to the schedule to address the funding crisis. The fact that there is now no one to turn to, she added, makes the current situation “not comparable to the past.”

But that’s not to say that everything was doom and gloom. “You could feel the solidarity and community,” says the EFF’s Cohn. “And having [the conference] in Taiwan, which lives in the shadow of a very powerful, often hostile government, seemed especially fitting.”

Indeed, if there was one theme that was repeated throughout the event, it was a shared desire to rethink and challenge who holds power.

Multiple sessions, for example, focused on strategies to counter both unresponsive Big Tech platforms and repressive governments. Meanwhile, during the session on AI and hate-speech moderation, participants concluded that one way of creating a safer internet would be for local organizations to build localized language models that are context- and language-specific. At the very least, said Curzi, we could move to other, smaller platforms that match our values, because at this point, “the big platforms can do anything they want.”

Do you have additional information on how Doge is affecting digital rights globally? Please use a non-work device and get in touch at [email protected] or with the reporter on Signal: eileenguo.15.

Stay Ahead

Explore More Insights

Stay ahead with more perspectives on cutting-edge power, infrastructure, energy, bitcoin and AI solutions. Explore these articles to uncover strategies and insights shaping the future of industries.

Gluware tackles AI agent coordination with Titan platform

The first phase focused on configuration management and drift detection. Gluware’s system identified when network devices deviated from approved configurations and proposed fixes, but network operations teams manually reviewed and approved each remediation. The second phase introduced automatic remediation. As customers gained confidence, they allowed the system to automatically correct

Ransomware gangs find a new hostage: Your AWS S3 buckets

To succeed, attackers typically look for S3 buckets that have: versioning disabled ( so old versions can’t be restored), object-lock disabled ( so files can be overwritten or deleted), wide write permissions (via mis-configured IAM policies or leaked credentials), and hold high-value data (backup files, production config dumps). Once inside,

Cisco initiative targets device security

Cisco is announcing a security initiative that will push customers to update or replace aging infrastructure components, such as routers, switches and firewalls, as well as discourage them from using any insecure features. Called Resilient Infrastructure, the plan calls for Cisco to strengthen network security by increasing default protections, removing

What are TPUs? Your guide to tensor processing units and AI acceleration

How TPUs work AI platforms and their underlying ML models require intensive mathematical processing. Fundamentally, TPUs are optimized for a type of mathematical operation known as tensor computation. Tensors are multi-dimensional arrays, or matrixes, that store and process data. Think of them as fundamental data structures or “gears” in machine

Reliance Stops Using Russian Oil in Part of Jamnagar

India’s Reliance Industries Ltd. said it would stop processing Russian oil at part of its giant Jamnagar oil refinery as US sanctions force the company to shy away from dealings with Moscow. The export-focused part of the refinery, which accounts for about half of its 1.4 million barrels a day of capacity, took its last shipment of Russian crude on Thursday, the company said in statement. The move would mean the site could keep supplying fuel to Europe when new sanctions banning the import of petroleum made from Russian crude come into effect early next year. It will also demonstrate compliance with a US effort to force processors away from Russian barrels. Reliance isn’t currently buying Russian oil and hasn’t taken a view yet on whether it will resume doing so, a person with knowledge of the matter said, asking not to be identified because the information isn’t public. Together, the two sites at Jamnagar make it the world’s biggest oil refinery. Still, the company said in a statement that some purchases bought before the US put sanctions on Russia’s two largest oil companies would discharge at another part of the Jamnagar facility that supplies the domestic market, it added. The US announcement of sanctions on Lukoil PJSC and Rosneft PJSC last month sent shockwaves through Asian oil buyers, as it meant a swath of Russia’s flows are pumped by blacklisted firms. Processors in India and China had snapped up cheap Russian barrels in the aftermath of the war in Ukraine, denting the impact of rampant global inflation in 2022. A deadline to wind down deals with the duo is set to pass on Friday, putting pressure on the companies and countries that had continued to buy barrels from Moscow after Russia invaded Ukraine. While Indian refiners have been booking

DOE Seeks Input on Gas Turbine Manufacturing to Increase Domestic Energy Production

WASHINGTON — The U.S. Department of Energy’s (DOE) Hydrocarbons and Geothermal Energy Office (HGEO) today issued a request for information (RFI) focused on evaluating challenges faced by U.S. manufacturers that currently constrain gas turbine production capacity. Natural gas turbines offer several benefits for domestic electricity generation, including high operational flexibility, efficiencies, and reliability. DOE will use stakeholder feedback to inform effective research and development that can increase the pace of manufacturing these crucial energy generating technologies. This effort supports President Trump’s commitment to boost production of our domestic energy resources to ensure affordable and reliable energy for all Americans and protect our national and economic security. According to the U.S. Energy Information Administration, electricity demand, which remained relatively flat for the last two decades, is now expected to grow at an average rate of 1.7% per year in the short-term forecast and to exceed 6,000 terawatt-hours by 2050, a 50% increase from 2024 levels. Such demand—due in part to rapid growth in data centers and artificial intelligence, reshoring of American manufacturing, and increased electrification in building operations, transportation, and industry—will strain equipment supply chains, especially gas turbines, which provide more than 40% of electricity in the United States. Further, with the recent spike in electricity demand, the delivery wait time for gas turbines has doubled from two to three years to as many as seven years, resulting in price increases due to the limited supply. Such long lead times and high prices threaten to constrain the effective supply of electricity to meet demand. To assist DOE in evaluating the gaps that constrain the production capacity of U.S. manufacturers of gas turbines, DOE is seeking input from interested parties in the categories of manufacturing technology, workforce, sub-suppliers, and materials. To review the RFI, please click here. Responses must be submitted electronically to [email protected], with the subject line “U.S.

Lukoil Dissolves International Board

Russian energy giant Lukoil PJSC dissolved the supervisory board of its international business, the latest sign of how US sanctions — the first of which begin on Friday — are affecting the firm. As part of the dissolution, the Moscow-based firm “recalled” Sergei Kochkurov, chief executive officer of the parent company, as well as Evgeny Khavkin and Gennady Fedotov. The step, taken during an Oct. 28 board meeting, was posted by Lukoil International GmbH on Austria’s corporate register on Friday. The US Treasury’s Office of Foreign Assets Control announced on Oct. 22 that it was sanctioning Lukoil and fellow Russian giant Rosneft PJSC. The measures start today although some actions against Lukoil assets have been delayed until Dec. 13. The move stressed the firm globally: Russian oil prices plunged, its international trading business Litasco has shed staff and wound up at least some operations. Lukoil’s share of revenue from the West Qurna 2 oil field in Iraq has been frozen by Baghdad and western suitors are circling the firm’s global assets. The decision to dissolve the board and recall Lukoil International’s overseers will leave the company’s managing director Alexander Matytsyn in charge. The company is still fully owned by Lukoil. On Wednesday, the Vienna-based unit also published its fully audited group report for 2022 — taking about two years longer than normal to do so. The move offered a first detailed view of how the company fared in the first year of Russia’s invasion of Ukraine. According to those accounts, completed by KPMG on Oct. 9 this year, Lukoil International booked €95 billion of revenue and a net income of €7.8 billion in 2022 — a period that reflected the height of the European energy crisis. Some of the world’s largest energy companies, including Exxon Mobil Corp., Chevron Corp. and

Texas loan fund tops 3.5 GW of gas capacity secured with latest NRG deal

NRG Energy will develop a 455-MW gas plant near Houston, backed by a low-interest loan from the state of Texas, the company said Thursday. It is the sixth loan finalized through the Texas Energy Fund program for the Electric Reliability Council of Texas market. Voters authorized the fund in 2023. The new generation will be built at NRG’s existing Greens Bayou Generating Station in Harris County, and is expected online in 2028. New generation backed by the Texas Energy Fund, across six projects, now exceeds more than 3.5 GW, said officials at the Public Utility Commission of Texas, which manages the program. The Lone Star State is experiencing “unprecedented growth,” NRG Executive Vice President, President of NRG Business and Wholesale Operations Robert Gaudette, said in a statement. Greens Bayou is NRG’s third project to receive support from the loan fund, marking about 1.5 GW of total capacity and up to $1.15 billion in low-interest loans. Under the loan agreement for Greens Bayou, total project costs are estimated to be less than $617 million and the PUCT will provide a 20-year loan up to $370 million, or 60% of total cost, at a 3% interest rate. In August, the PUCT tapped NRG for a TEF loan up to $216 million to develop two gas units totaling 456 MW of capacity at its existing TH Wharton Generating Station in Houston. And in September the company was selected for a loan of $562 million to develop a 721-MW gas plant near Baytown. There are another 11 Texas Energy Fund applications moving through a due diligence review process, said state officials, representing another 5.4 GW of possible new capacity in the ERCOT footprint.

The week in 5 numbers: Power outages stretch, data center load overstated

Capacity secured by the Texas Energy Fund so far for the Electric Reliability Council of Texas market. The fund, which voters authorized in 2023 to give low-interest loans to energy projects, was a response to the devastating impacts of 2021’s winter storm blackouts. It finalized its sixth loan this week, which went to NRG Energy to develop a 455-MW gas plant near Houston. The new generation will be built at NRG’s existing Greens Bayou Generating Station in Harris County, and it is expected to come online in 2028. The fund has faced challenges, however, as a number of projects have withdrawn from the pipeline as the economics of gas plants, in particular, have shifted.

StanChart Flags ‘Fragility of Russian Supply’

In a market that continues to be dominated by bearish supply sentiment, one of the few bullish drivers is the fragility of Russian supply. That’s what Standard Chartered Bank Energy Research Head Emily Ashford said in a report sent to Rigzone by the Standard Chartered team earlier this week, adding that both crude and refined product exports are being curtailed by the pressure of focused sanctions on Russian oil producers, a lowered crude oil price cap, and ongoing missile and drone attacks on oil and gas export infrastructure. “On 14 November the port of Novorossiysk, in Krasnodar Krai, was targeted by missiles and drones, with a focus on the Sheskharis oil terminal,” Ashford highlighted in the report, noting that the terminal “has an export capacity of c.2.2 million barrels per day” and that “loadings were suspended for two days”. “Ukraine’s attacks on a series of Black Sea export terminals have highlighted the vulnerability of exports via Russia’s southern route,” Ashford noted in the report. “This is particularly important, with weather closing down the Northern Sea Route via the Arctic over the winter. The winter transit routes to Asia are then limited to the Suez Canal and take, on average, 10 days longer,” Ashford added. “These longer transits are a contributing factor to the increased volumes of oil on water, which have increased by 294 million barrels year on year to an all-time high of 1.37 billion barrels as of 14 November, according to data from Vortexa,” Ashford continued. In the report, Ashford went on to note that Russian crude exports “have … remained relatively steady” but added that Standard Chartered expects to see “a sharp slowdown after the 21 November deadline for dealings with the two sanctioned oil producers”. “This is likely also a contributing factor to the volumes of

Microsoft’s Fairwater Atlanta and the Rise of the Distributed AI Supercomputer

Microsoft’s second Fairwater data center in Atlanta isn’t just “another big GPU shed.” It represents the other half of a deliberate architectural experiment: proving that two massive AI campuses, separated by roughly 700 miles, can operate as one coherent, distributed supercomputer. The Atlanta installation is the latest expression of Microsoft’s AI-first data center design: purpose-built for training and serving frontier models rather than supporting mixed cloud workloads. It links directly to the original Fairwater campus in Wisconsin, as well as to earlier generations of Azure AI supercomputers, through a dedicated AI WAN backbone that Microsoft describes as the foundation of a “planet-scale AI superfactory.” Inside a Fairwater Site: Preparing for Multi-Site Distribution Efficient multi-site training only works if each individual site behaves as a clean, well-structured unit. Microsoft’s intra-site design is deliberately simplified so that cross-site coordination has a predictable abstraction boundary—essential for treating multiple campuses as one distributed AI system. Each Fairwater installation presents itself as a single, flat, high-regularity cluster: Up to 72 NVIDIA Blackwell GPUs per rack, using GB200 NVL72 rack-scale systems. NVLink provides the ultra-low-latency, high-bandwidth scale-up fabric within the rack, while the Spectrum-X Ethernet stack handles scale-out. Each rack delivers roughly 1.8 TB/s of GPU-to-GPU bandwidth and exposes a multi-terabyte pooled memory space addressable via NVLink—critical for large-model sharding, activation checkpointing, and parallelism strategies. Racks feed into a two-tier Ethernet scale-out network offering 800 Gbps GPU-to-GPU connectivity with very low hop counts, engineered to scale to hundreds of thousands of GPUs without encountering the classic port-count and topology constraints of traditional Clos fabrics. Microsoft confirms that the fabric relies heavily on: SONiC-based switching and a broad commodity Ethernet ecosystem to avoid vendor lock-in and accelerate architectural iteration. Custom network optimizations, such as packet trimming, packet spray, high-frequency telemetry, and advanced congestion-control mechanisms, to prevent collective

Land & Expand: Hyperscale, AI Factory, Megascale

Land & Expand is Data Center Frontier’s periodic roundup of notable North American data center development activity, tracking the newest sites, land plays, retrofits, and hyperscale campus expansions shaping the industry’s build cycle. October delivered a steady cadence of announcements, with several megascale projects advancing from concept to commitment. The month was defined by continued momentum in OpenAI and Oracle’s Stargate initiative (now spanning multiple U.S. regions) as well as major new investments from Google, Meta, DataBank, and emerging AI cloud players accelerating high-density reuse strategies. The result is a clearer picture of how the next wave of AI-first infrastructure is taking shape across the country. Google Begins $4B West Memphis Hyperscale Buildout Google formally broke ground on its $4 billion hyperscale campus in West Memphis, Arkansas, marking the company’s first data center in the state and the anchor for a new Mid-South operational hub. The project spans just over 1,000 acres, with initial site preparation and utility coordination already underway. Google and Entergy Arkansas confirmed a 600 MW solar generation partnership, structured to add dedicated renewable supply to the regional grid. As part of the launch, Google announced a $25 million Energy Impact Fund for local community affordability programs and energy-resilience improvements—an unusually early community-benefit commitment for a first-phase hyperscale project. Cooling specifics have not yet been made public. Water sourcing—whether reclaimed, potable, or hybrid seasonal mode—remains under review, as the company finalizes environmental permits. Public filings reference a large-scale onsite water treatment facility, similar to Google’s deployments in The Dalles and Council Bluffs. Local governance documents show that prior to the October announcement, West Memphis approved a 30-year PILOT via Groot LLC (Google’s land assembly entity), with early filings referencing a typical placeholder of ~50 direct jobs. At launch, officials emphasized hundreds of full-time operations roles and thousands

The New Digital Infrastructure Geography: Green Street’s David Guarino on AI Demand, Power Scarcity, and the Next Phase of Data Center Growth

As the global data center industry races through its most frenetic build cycle in history, one question continues to define the market’s mood: is this the peak of an AI-fueled supercycle, or the beginning of a structurally different era for digital infrastructure? For Green Street Managing Director and Head of Global Data Center and Tower Research David Guarino, the answer—based firmly on observable fundamentals—is increasingly clear. Demand remains blisteringly strong. Capital appetite is deepening. And the very definition of a “data center market” is shifting beneath the industry’s feet. In a wide-ranging discussion with Data Center Frontier, Guarino outlined why data centers continue to stand out in the commercial real estate landscape, how AI is reshaping underwriting and development models, why behind-the-meter power is quietly reorganizing the U.S. map, and what Green Street sees ahead for rents, REITs, and the next wave of hyperscale expansion. A ‘Safe’ Asset in an Uncertain CRE Landscape Among institutional investors, the post-COVID era was the moment data centers stepped decisively out of “niche” territory. Guarino notes that pandemic-era reliance on digital services crystallized a structural recognition: data centers deliver stable, predictable cash flows, anchored by the highest-credit tenants in global real estate. Hyperscalers today dominate new leasing and routinely sign 15-year (or longer) contracts, a duration largely unmatched across CRE categories. When compared with one-year apartment leases, five-year office leases, or mall anchor terms, the stability story becomes plain. “These are AAA-caliber companies signing the longest leases in the sector’s history,” Guarino said. “From a real estate point of view, that combination of tenant quality and lease duration continues to position the asset class as uniquely durable.” And development returns remain exceptional. Even without assuming endless AI growth, the math works: strong demand, rising rents, and high-credit tenants create unusually predictable performance relative to

The Flexential Blueprint: New CEO Ryan Mallory on Power, AI, and Bending the Physics Curve

In a coordinated leadership transition this fall, Ryan Mallory has stepped into the role of CEO at Flexential, succeeding Chris Downie. The move, described as thoughtful and planned, signals not a shift in direction, but a reinforcement of the company’s core strategy, with a sharpened focus on the unprecedented opportunities presented by the artificial intelligence revolution. In an exclusive interview on the Data Center Frontier Show Podcast, Mallory outlined a confident vision for Flexential, positioning the company at the critical intersection of enterprise IT and next-generation AI infrastructure. “Flexential will continue to focus on being an industry and market leader in wholesale, multi-tenant, and interconnection capabilities,” Mallory stated, affirming the company’s foundational strengths. His central thesis is that the AI infrastructure boom is not a monolithic wave, but a multi-stage evolution where Flexential’s model is uniquely suited for the emerging “inference edge.” The AI Build Cycle: A Three-Act Play Mallory frames the AI infrastructure market as a three-stage process, each lasting roughly four years. We are currently at the tail end of Stage 1, which began with the ChatGPT explosion three years ago. This phase, characterized by a frantic rush for capacity, has led to elongated lead times for critical infrastructure like generators, switchgear, and GPUs. The capacity from this initial build-out is expected to come online between late 2025 and late 2026. Stage 2, beginning around 2026 and stretching to 2030, will see the next wave of builds, with significant capacity hitting the market in 2028-2029. “This stage will reveal the viability of AI and actual consumption models,” Mallory notes, adding that air-cooled infrastructure will still dominate during this period. Stage 3, looking ahead to the early 2030s, will focus on long-term scale, mirroring the evolution of the public cloud. For Mallory, the enduring nature of this build cycle—contrasted

Centersquare Launches $1 Billion Expansion to Scale an AI-Ready North American Data Center Platform

A Platform Built for Both Colo and AI Density The combined Evoque–Cyxtera platform entered the market with hundreds of megawatts of installed capacity and a clear runway for expansion. That scale positioned Centersquare to offer both traditional enterprise colocation and the higher-density, AI-ready footprints increasingly demanded through 2024 and 2025. The addition of these ten facilities demonstrates that the consolidation strategy is gaining traction, giving the platform more owned capacity to densify and more regional optionality as AI deployment accelerates. What’s in the $1 Billion Package — and Why It Matters 1) Lease-to-Own Conversions in Boston & Minneapolis Centersquare’s decision to purchase two long-operated but previously leased sites in Boston and Minneapolis reduces long-term occupancy risk and gives the operator full capex control. Owning the buildings unlocks the ability to schedule power and cooling upgrades on Centersquare’s terms, accelerate retrofits for high-density AI aisles, deploy liquid-ready thermal topologies, and add incremental power blocks without navigating landlord approval cycles. This structural flexibility aligns directly with the platform’s “AI-era backbone” positioning. 2) Eight Additional Data Centers Across Six Metros The acquisitions broaden scale in fast-rising secondary markets—Tulsa, Nashville, Raleigh—while deepening Centersquare’s presence in Dallas and expanding its Canadian footprint in Toronto and Montréal. Dallas remains a core scaling hub, but Nashville and Raleigh are increasingly important for enterprises modernizing their stacks and deploying regional AI workloads at lower cost and with faster timelines than congested Tier-1 corridors. Tulsa provides a network-adjacent, cost-efficient option for disaster recovery, edge aggregation, and latency-tolerant compute. In Canada, Toronto and Montréal offer strong enterprise demand, attractive economics, and grid advantages—including Québec’s hydro-powered, low-carbon energy mix—that position them well for AI training spillover and inference workloads requiring reliable, competitively priced power. 3) Self-Funded With Cash on Hand In the current rate environment, funding the entire $1 billion package

Fission Forward: Next-Gen Nuclear Power Developments for the AI Data Center Boom

Constellation proposes to begin with 1.5 GW of fast-tracked projects, including 800 MW of battery energy storage and 700 MW of new natural gas generation to address short-term reliability needs. The remaining 4.3 GW represents longer-term investment at the Calvert Cliffs Clean Energy Center: extending both units for an additional 20 years beyond their current 2034 and 2036 license expirations, implementing a 10% uprate that would add roughly 190 MW of output, and pursuing 2 GW of next-generation nuclear at the existing site. For Maryland, a state defined by a dense I-95 fiber corridor, accelerating data center buildout, and rising AI-driven load, the plan could be transformative. If Constellation moves from “option” to “program,” the company estimates that 70% of the state’s electricity supply could come from clean energy sources, positioning Maryland as a top-tier market for 24/7 carbon-free power. TerraPower’s Natrium SMR Clears a Key Federal Milestone On Oct. 23, the Nuclear Regulatory Commission issued the final environmental impact statement (FEIS) for TerraPower’s Natrium small modular reactor in Kemmerer, Wyoming. While not a construction permit, FEIS completion removes a major element of federal environmental risk and keeps the project on track for the next phase of NRC review. TerraPower and its subsidiary, US SFR Owner, LLC, originally submitted the construction permit application on March 28, 2024. Natrium is a sodium-cooled fast reactor producing roughly 345 MW of electric output, paired with a molten-salt thermal-storage system capable of boosting generation to about 500 MW during peak periods. The design combines firm baseload power with flexible, dispatchable capability, an attractive profile for hyperscalers evaluating 24/7 clean energy options in the western U.S. The project is part of the DOE’s Advanced Reactor Demonstration Program, intended to replace retiring coal capacity in PacifiCorp’s service territory while showcasing advanced fission technology. For operators planning multi-GW

Microsoft will invest $80B in AI data centers in fiscal 2025

And Microsoft isn’t the only one that is ramping up its investments into AI-enabled data centers. Rival cloud service providers are all investing in either upgrading or opening new data centers to capture a larger chunk of business from developers and users of large language models (LLMs). In a report published in October 2024, Bloomberg Intelligence estimated that demand for generative AI would push Microsoft, AWS, Google, Oracle, Meta, and Apple would between them devote $200 billion to capex in 2025, up from $110 billion in 2023. Microsoft is one of the biggest spenders, followed closely by Google and AWS, Bloomberg Intelligence said. Its estimate of Microsoft’s capital spending on AI, at $62.4 billion for calendar 2025, is lower than Smith’s claim that the company will invest $80 billion in the fiscal year to June 30, 2025. Both figures, though, are way higher than Microsoft’s 2020 capital expenditure of “just” $17.6 billion. The majority of the increased spending is tied to cloud services and the expansion of AI infrastructure needed to provide compute capacity for OpenAI workloads. Separately, last October Amazon CEO Andy Jassy said his company planned total capex spend of $75 billion in 2024 and even more in 2025, with much of it going to AWS, its cloud computing division.

John Deere unveils more autonomous farm machines to address skill labor shortage

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Self-driving tractors might be the path to self-driving cars. John Deere has revealed a new line of autonomous machines and tech across agriculture, construction and commercial landscaping. The Moline, Illinois-based John Deere has been in business for 187 years, yet it’s been a regular as a non-tech company showing off technology at the big tech trade show in Las Vegas and is back at CES 2025 with more autonomous tractors and other vehicles. This is not something we usually cover, but John Deere has a lot of data that is interesting in the big picture of tech. The message from the company is that there aren’t enough skilled farm laborers to do the work that its customers need. It’s been a challenge for most of the last two decades, said Jahmy Hindman, CTO at John Deere, in a briefing. Much of the tech will come this fall and after that. He noted that the average farmer in the U.S. is over 58 and works 12 to 18 hours a day to grow food for us. And he said the American Farm Bureau Federation estimates there are roughly 2.4 million farm jobs that need to be filled annually; and the agricultural work force continues to shrink. (This is my hint to the anti-immigration crowd). John Deere’s autonomous 9RX Tractor. Farmers can oversee it using an app. While each of these industries experiences their own set of challenges, a commonality across all is skilled labor availability. In construction, about 80% percent of contractors struggle to find skilled labor. And in commercial landscaping, 86% of landscaping business owners can’t find labor to fill open positions, he said. “They have to figure out how to do

2025 playbook for enterprise AI success, from agents to evals

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More 2025 is poised to be a pivotal year for enterprise AI. The past year has seen rapid innovation, and this year will see the same. This has made it more critical than ever to revisit your AI strategy to stay competitive and create value for your customers. From scaling AI agents to optimizing costs, here are the five critical areas enterprises should prioritize for their AI strategy this year. 1. Agents: the next generation of automation AI agents are no longer theoretical. In 2025, they’re indispensable tools for enterprises looking to streamline operations and enhance customer interactions. Unlike traditional software, agents powered by large language models (LLMs) can make nuanced decisions, navigate complex multi-step tasks, and integrate seamlessly with tools and APIs. At the start of 2024, agents were not ready for prime time, making frustrating mistakes like hallucinating URLs. They started getting better as frontier large language models themselves improved. “Let me put it this way,” said Sam Witteveen, cofounder of Red Dragon, a company that develops agents for companies, and that recently reviewed the 48 agents it built last year. “Interestingly, the ones that we built at the start of the year, a lot of those worked way better at the end of the year just because the models got better.” Witteveen shared this in the video podcast we filmed to discuss these five big trends in detail. Models are getting better and hallucinating less, and they’re also being trained to do agentic tasks. Another feature that the model providers are researching is a way to use the LLM as a judge, and as models get cheaper (something we’ll cover below), companies can use three or more models to

OpenAI’s red teaming innovations define new essentials for security leaders in the AI era

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More OpenAI has taken a more aggressive approach to red teaming than its AI competitors, demonstrating its security teams’ advanced capabilities in two areas: multi-step reinforcement and external red teaming. OpenAI recently released two papers that set a new competitive standard for improving the quality, reliability and safety of AI models in these two techniques and more. The first paper, “OpenAI’s Approach to External Red Teaming for AI Models and Systems,” reports that specialized teams outside the company have proven effective in uncovering vulnerabilities that might otherwise have made it into a released model because in-house testing techniques may have missed them. In the second paper, “Diverse and Effective Red Teaming with Auto-Generated Rewards and Multi-Step Reinforcement Learning,” OpenAI introduces an automated framework that relies on iterative reinforcement learning to generate a broad spectrum of novel, wide-ranging attacks. Going all-in on red teaming pays practical, competitive dividends It’s encouraging to see competitive intensity in red teaming growing among AI companies. When Anthropic released its AI red team guidelines in June of last year, it joined AI providers including Google, Microsoft, Nvidia, OpenAI, and even the U.S.’s National Institute of Standards and Technology (NIST), which all had released red teaming frameworks. Investing heavily in red teaming yields tangible benefits for security leaders in any organization. OpenAI’s paper on external red teaming provides a detailed analysis of how the company strives to create specialized external teams that include cybersecurity and subject matter experts. The goal is to see if knowledgeable external teams can defeat models’ security perimeters and find gaps in their security, biases and controls that prompt-based testing couldn’t find. What makes OpenAI’s recent papers noteworthy is how well they define using human-in-the-middle

Stay Ahead, Stay ONMINE