Image of map and computer language over binary codes on server racks in server room.

Global Digital Policy Roundup: August 2026

Electrical infrastructure with a data center in the background, and a large puddle of water on the ground with a reflection of the infrastructure. 

Electrical infrastructure at a Flexential data center in Hillsboro, Oregon, on May 16, 2026. (Photo by Alex Milan Tracy/Sipa USA)

To Change Data Centers, Flip the Narrative on Sustainability

A major data center project underway in Cape Town, South Africa. Photo: David Harrison/The Continent 

Cape Town is Approving Data Centers with Rules That Still Treat Them Like Warehouses

Students return to classrooms in Tuxtla Gutiérrez as the 2026–2027 school year begins across Chiapas. (Photo by Diego Pérez Gómez / JNA Press/Sipa via AP Images)

Building LLMs? Think Beyond Borders

Parents and teachers held a rally calling on Mayor Mamdani and NYC School Chancellor Kamar Samuels to have a two year moratorium on AI in public schools in New York City, NY. July 29, 2026 (Photo by Steve Sanchez/Sipa via AP Images)

Across the US, School Districts Wrestle With Classroom AI Policy

Deepfake vector illustration. Fake media method in tiny persons concept. Use artificial neural networks to make hoaxes video. Falsified machine learning autoencoders system. Fraud systems development.

China’s Disinformation Operations in Taiwan Are Changing with AI

New York Gov. Kathy Hochul (center), flanked by Attorney General Letitia James (right) and supporters, holds up one of two bills she signed in July 2024 aimed at protecting children and youth from the harmful effects of social media. Photo by Susan Watts/Office of New York Gov. Source

New York Gov. Kathy Hochul (center), flanked by Attorney General Letitia James (right) and supporters, holds up one of two bills she signed in July 2024 aimed at protecting children and youth from the harmful effects of social media. Photo by Susan Watts/Office of New York Gov. 

As Age Assurance Moves Into Practice, What Can Policymakers Learn From New York?

An illustration to depict a conceptual network, with various points connected by strings of different colors. The visualization represents networks, relationships, or interconnected systems, such as in social media, data flow, or neural networks.

Why the UK Needs Better Social Media Data Access to Research Online Harms

Within the frames are people bound to their office cubicles; beyond them, individuals work freely from diverse locations, connected through digital signals.

Digital Nomads Beyond the Cubicle by Yutong Liu & Digit / 

This Labor Day, Consider AI's Impact on Hiring

United States Representative Zoe Lofgren (D-Calif.) speaking at a hearing of the House Committee on Science, Space, and Technology Subcommittee on Energy at the US Capitol in Washington, DC, on September 2, 2026. (Photo by Michael Brochstein/Sipa via AP Images)

Rep. Zoe Lofgren on FISA, Surveillance and the Fourth Amendment

The landmark front entrance of the Virginia State Capitol in Richmond, Virginia, the historic seat of the commonwealth government, on April 18, 2025. Shutterstock

The landmark front entrance of the Virginia State Capitol in Richmond, Virginia, the historic seat of the commonwealth government, on April 18, 2025. 

Virginia Eyes AI Chatbot Rules as Observable Risks Warrant Action

New York City Schools Chancellor Kamar Samuels and Mayor Zohran Mamdani hold a press conference at the Brooklyn STEAM Center on September 2, 2026 in New York. Photo by: NDZ/STAR MAX/IPx

America’s Two Largest School Districts Impose AI Moratoriums

The OpenAI logo is displayed on a mobile phone screen in this photo illustration, as artificial intelligence tools continue to expand across consumer and business applications in Brussels, Belgium, on February 28, 2026. (Photo by Jonathan Raa/NurPhoto via AP)

What ChatGPT’s DSA Designation Means for OpenAI and the EU

Google CEO Sundar Pichai talks before a dinner with President Donald Trump and Japan's Prime Minister Sanae Takaichi in the State Dining Room of the White House, Thursday, March 19, 2026, in Washington. (AP Photo/Julia Demaree Nikhinson)

The Google Ad Tech Remedies Matter to You, Too

The Meta logo on a smartphone against a red background.

Without a Child Rights Framework, Meta’s $17.1bn Settlement Will Only Go So Far

Demonstrators take part in a protest against the deployment of an AI-assisted police surveillance system in Berlin on July 17, 2026. Photo by Andres Trujilo/ JNA PRESS(Sipa via AP Images)

Violent Anti-Tech Extremism Has Roots in History

Illustration of a laptop displaying the OpenAI logo on its screen. A hand holds a magnifying glass over the screen with blue lightning bolts emanating from it, symbolizing scrutiny or investigation. The image uses a halftone comic style with blue and black accents.

ChatGPT, Roblox Face Tough EU Rules. Implementation Is the Hard Part

Paper collage. A photograph of the Atomium in Brussels, cut into diagonal strips and reassembled over a grid of numbers.

Seeing More — Seeing Less by Anna Riepe & FARI / 

Can the EU AI Act’s Transparency Rules Strengthen Democratic Resilience?

﻿Germany’s Federal Chancellery building on a sunny day

Germany’s Federal Chancellery building on July 19, 2025 (Photo by Michael Nguyen/NurPhoto) 

What Germany's AI Security Institute Could Learn From Other Countries

This image is a section of the Detroit Industry Murals by Diego Rivera, created between 1932 and 1933. Rivera, a renowned Mexican muralist, painted these panels inside the Detroit Institute of Arts. They are widely considered one of his greatest works and a landmark of 20th-century art.

The workers shown are laboring on automobile assembly lines, likely at the Ford Motor Company’s River Rouge Plant.

Rivera celebrated the dignity and effort of industrial labor, portraying the workers with strength and unity.

The style is Social Realism, characterized by its focus on working-class subjects and its detailed, lifelike forms.

Photograph of a detail of a 1933 Diego Rivera fresco at the Detroit Institute of Arts. Justin Hendrix/Tech Policy Press

A Labor Day Prescription for Preserving Fair Employment in the Digital Age

Students return to classrooms in Tuxtla Gutiérrez as the 2026–2027 school year begins across Chiapas. (Photo by Diego Pérez Gómez / JNA Press/Sipa via AP Images)


In early 2023, just following the public launch of ChatGPT, discussion surrounding the opportunities and risks of using LLMs across a variety of disciplines exploded. Legal professionals, educators, and journalists were among those beginning to explore how they could use LLMs in their work. With little practical research and few recommendations available across the fields of law, education, and journalism at the time, we began an 

analysis of the human rights risks and opportunities posed by the use of LLMs

 in these three fields. One of the biggest lessons we learned was the need to engage with users all around the world, because the way practitioners use LLMs in their work varies significantly based on where they live and work. While there were some exceptions, we found that practitioners in Global South states tended to use LLMs more ambitiously in their work, while those in Global North states were typically more risk-averse.

To learn more, we interviewed 56 experts and professionals across 24 countries in seven regions of the world to find out how they use LLMs in their work. We found that using LLMs for more complex, high-value tasks, as many did in the Global South, can expand human rights opportunities for people but can also heighten the likelihood and severity of risks people may endure. Where institutional capacity is already stretched thin, LLMs have started to fill gaps, substituting for lawyers, teachers, and journalists.

The critical gaps that LLMs are used to fill contribute to the generally positive attitudes toward LLMs expressed by the people we interviewed in South America, Sub-Saharan Africa, Asia, the Middle East, and North Africa, who describe LLMs as having positive effects in their communities. This contrasts with the experiences shared with us by many of their more risk-averse counterparts in Europe, North America, and Oceania. The optimism of experts and practitioners based in the Global South, however, is tempered: interviewees describe how uneven access to devices, training, and connectivity in their communities forcibly limits those who can use LLMs at all and how well the tools perform for required tasks. In this piece, we share stories told to us anonymously by people we interviewed in the Global South to illustrate how LLMs create opportunities, even as they introduce risks.

In Singapore, LLMs are used in small claims tribunals to provide legal information to litigants. Because no lawyers are involved in such tribunals, the idea is to help civilians who are unclear about how to file legal documents. This means that a conversational chatbot provides access to this information without legal fees. This has helped democratize access to justice, providing more assistance to litigants than they may otherwise have been able to afford. On the flip side, if the chatbot provides incorrect information, or an attorney could have provided additional information to further help a claimant’s chances for success, this could risk impacting their right to access justice. Interviewees in both Singapore and Qatar described how LLMs are being used by legal professionals to translate interviews and other documents to or from English or between other languages. While this helps advance the speediness of accessing justice, it also risks harming it if gaps exist in the translation of the document because the LLM lacks certain language capabilities. It also requires that legal professionals double check the translation or risk inaccuracy.

In South Africa, journalists use LLMs for research and writing stories that tend to be more formulaic, such as those based on press releases or sporting events. This means that journalists there can focus on tasks that require critical thinking and emotional intelligence, leaving repetitive tasks to machines. For instance, rather than spending time writing about the outcome of a cricket match, a journalist would have the time to research and write an in-depth feature of a female cricket team. Journalists also rely on LLMs for sports reporting in India, as well as weather reports and business stories on market shares. For example, instead of writing a weather report for the forthcoming week, a journalist could focus on the upcoming monsoon season, detailing its likely severity and steps people can take to stay safe. This makes it easier for journalistic organizations to provide access to both routine and analytical information to readers and listeners. However, it also means that if journalists or editors do not double check the content, inaccurate information could be provided to readers and listeners who then rely upon it for making decisions.

In the United Arab Emirates, an interviewee described an AI tutor providing personalized learning to thousands of students. The tool supports the right to education for students who struggle with the language of instruction by translating lessons between English and Arabic, and for students with disabilities by transforming educational content to modalities like audio. In Mexico, educators with access and exposure to LLMs have used them innovatively in classrooms. In one instance, a teacher asked students to use ChatGPT to answer math and physics questions relevant to the curriculum and then required students to assess its responses for accuracy. Such efforts have sought to strengthen critical thinking in AI use, which is essential to protecting freedom of thought.

However, Mexico faces stark disparities in access and training, with schools in some regions having access to only a single computer for the entire building and some educators reporting that they have received no guidance about how to use LLMs in their work. While LLMs may provide opportunities to students, guidance needs to be provided to teachers and students about how to use LLMs effectively. When access is not available, such opportunities fail to reach some students, which risks harming those students’ right to education. Without clear guidance to students, they may rely more heavily on LLMs than on their own critical thinking, potentially impacting their freedom of thought.

If it occurs at all, most assessment of human rights risks by LLM companies happens at the deployment stage, where models are integrated into user-facing applications. And while these assessments sometimes focus on outputs in countries outside of the United States, no company we spoke with told us they had evaluated harms in numerous countries at once. In all of our interviews with LLM companies, representatives described engineering practices to evaluate and align models in accordance with values specified in their responsible AI policies—which aligned explicitly with human rights norms and instruments only in some cases. But model developers and deployers must do more.

Despite growing awareness of human rights risks posed by LLMs, most model-level evaluations remain focused on metrics for general performance, and sometimes specific risks like social biases, without explicitly linking those metrics to human rights outcomes. Model evaluations can be intentionally designed around human rights impacts. In 

 focused on the use of LLMs in political news journalism, we propose and demonstrate a rights-aligned evaluation framework that draws on human rights instruments to center rights from the outset and link model behavior to real-world consequences. We recommend aligning evaluation tasks with specific human rights risks; developing metrics that capture the risks and their frequency; and interpreting results based on the seriousness, likelihood, and scope of the impact as well as the degree to which it is possible to remediate the impact.

Nonetheless, engineering is not enough. Model benchmarks do not capture the full context. Human rights stakeholder engagement is precisely designed to consider how geopolitics, resource and access disparities, and other factors shape how people use LLMs, and therefore the risks and opportunities they pose. LLM developers and deployers need to explore how people around the world use their products. Benchmarks measure what models can do in theory, but only people can describe their impacts in practice.

Electrical infrastructure at a data center facility in Hillsboro, Oregon with a puddle of water in the foreground, reflecting the building upside down. 

The AI race has spurred a worldwide frenzy to build massive new data centers, especially in the United States. Data centers 

 to consume up to 12% of US electricity by 2030, and their combined water usage could increase 

In response, a nationwide community backlash has surged. Between May 2024 and March 2026, 

 worth of new projects were canceled or stalled due to local opposition; $130 billion of that occurred in the first quarter of 2026 alone. 

 Americans now oppose a data center being built near them. Even 

, the famed 1990s environmental activist, is back on the scene with a 

 crowdsourcing community complaints; it has already received over 9,000 submissions.

But is the backlash actually prompting data center developers to change their practices?

 conference in New York to get a firsthand sense of how data center developers were perceiving the backlash.

Perhaps as a threat? In the breakfast line at the conference, a man next to me noted that one attendee had brought a personal body guard, referencing the recent (albeit unrelated) shootings in New York City of executives at Blackstone and UnitedHealthcare.

The people I spoke to on the exhibition floor, where thousands of people (

) mingled, viewed sustainability as a nice-to-have rather than as an imperative. On stage, many panelists discussed the importance of energy efficiency, better transparency, and community engagement. A select few presented on sustainability initiatives, such as 

Even so, what I heard expressed most, was frustration at what they see as the unfair demonization of an industry that has been around for decades and provides the backbone of the internet.

What I heard then, and from other data center professionals since, is that the industry’s priorities are as follows: speed, speed, speed, cost, and only then sustainability.

For those of us concerned about the sustainability impacts of the data center buildout, the need for solutions feels urgent. In the United States, the present possibility of a nationwide policy strategy to reduce data centers’ environmental footprint is 

While local regulators have some leverage, we must also find ways to incentivize data center developers to adopt sustainable practices voluntarily. This doesn’t preclude a larger conversation about the 

 of the rapid data center buildout, but I do believe their interests (higher financial returns) can align with my interests (positive impact on the climate and environment). If they must build, they should build greener.

Let’s consider the many ways in which sustainability is good for business:

Reduced construction times and operational costs

Data centers are expensive to build, requiring a lot of upfront capital costs. The frontier AI companies 

have already spent over a trillion on investments

 while generating only a fraction in revenue. The backlash is causing data center costs to increase, as local tax breaks are repealed, projects face delays, and facilities cover higher utility costs. Geopolitical tensions, including the US-Iran war, are pushing costs in the AI supply chain up further.

Meanwhile, clean technologies have become more cost competitive.

Given that a typical data center’s highest operating cost is energy, energy efficiency should be a top priority. Afterall, the cheapest megawatt is the one you don’t have to build. In one 

, reducing Power Usage Effectiveness (the ratio between total energy consumed and energy consumed from IT equipment) by 16% saved $262,000 annually in a 1 MW data center.

Data centers can procure renewable energy, which remains the 

 form of new-build energy generation, even without tax subsidies. For data center operators concerned about the intermittent nature of solar and wind power, significant 

 in battery storage make batteries a clean and reliable addition to the energy mix.

By contrast, gas-fired power plants are getting more expensive to build. Recent 

 from BloombergNEF shows that costs have risen 66% since 2023. The average construction time now takes 23% longer, with current waitlists for new gas turbines extending beyond 2028.

Water usage and noise pollution are among the most contentious issues in local disputes over data center development. Advanced liquid cooling methods, such as direct-to-chip cooling, use less 

 and improve energy efficiency, helping reduce operating costs. They are also quieter because they require less noisy, high-speed fans.

 into their design and construction. Reusing structural materials can be cheaper than sourcing virgin materials, reducing costs by 

 over a building’s lifecycle. Retired equipment can be repurposed and resold. Currently, only 

 from data center e-waste is recycled and reused—a missed business opportunity.

Moreover, repurposing vacant industrial buildings, or even 

, would reduce the need to clear land for new facilities. The added benefit is that many of these sites are already connected to energy grids and water infrastructure, helping to expedite permitting and construction times.

Better risk mitigation and future-proofing

Circularity also reduces supply chain risks. Critical minerals like copper, lithium, and aluminum have few substitutes and are resourced through highly concentrated supply chains. The combination of increasing demand and 

 has prompted new startups to focus on recovering critical minerals from wastestreams.

By investing in sustainability, data centers reduce another form of risk: compliance risk. Last year, 

 were introduced across all 50 states to regulate data centers. That regulatory momentum has continued in 2026, with local 

, and statewide pauses on new large data centers in 

Prioritizing sustainability positions facilities as “future-proof” assets that will meet tomorrow’s regulations today. This proactive approach not only increases the likelihood of getting projects greenlit, but potentially prevents costly retrofits, avoids regulatory penalties, and increases the probability that facilities can operate without interruption as standards evolve.

Enhanced public trust and brand reputation

Data center developers who prioritize sustainability, transparency, and genuine community engagement, from planning to operation, will differentiate themselves and increase stakeholder trust.

Companies that disclose their environmental impacts yield higher financial returns, with companies in emissions-intensive industries such as energy, mining, and manufacturing demonstrating the 

. Sustainability also enhances B2B competitiveness, as lower-emission data centers are better positioned to help customers meet their own 

The unprecedented scale of data center growth also creates unique opportunities to drive innovation on sustainability. Large data center operators can serve as anchor customers for emerging technologies, helping de-risk new energy and water technologies.

One example includes working with municipal leaders to feed data center waste heat into district heating systems, greenhouses, or aquaculture systems. 

 is also prompting AI data centers to incorporate flexibility into their electricity usage, especially during periods of peak demand.

Policymakers should offer both carrots and sticks

While market incentives can raise the ceiling on data center sustainability, regulations should still be enacted to raise the floor, ensuring no active harm is done.

The first priority should be increasing transparency, so that more stakeholders can gain a complete understanding of data center impacts based on auditable measurements.

This should begin during the permitting process, with developers conducting site assessments of expected impacts, with third party verification. At minimum, data centers should be required to publicly disclose their water and energy use. Water disclosures should extend beyond onsite usage to include water consumed at the power source.

During emergencies or peak demand, such as heat waves, municipalities should reserve the right to curtail data center power use to avoid straining the grid. This can go hand-in-hand with offering faster interconnection times to energy-efficient facilities, like in 

, where data centers with flexible connection agreements (FCAs) get faster grid connections.

Tax incentives should be reserved for developers that can credibly demonstrate commitments to sustainable practices. In the UK, for example, data centers can claim 

on investments in energy saving technology.

Finally, government officials should use the power of the purse through their own procurement practices and only award IT contracts to tech companies that meet specific criteria.

The conversation around data centers primarily centers on 

this infrastructure is built. Less attention is given to 

. Back at the DCD conference, I kept hearing the refrain that we need to “win the AI race” with vague allusions to “beating China.”

Yet, in the US, people are not only concerned about data centers. They are becoming 

 about AI itself, questioning whether this so-called race will leave them behind. Without good answers to questions that extend beyond the environmental impacts to jobs and other AI risks, data centers will continue to face opposition.

Perhaps it’s time to reframe the “AI race”, not as merely a geopolitical one, but as a race against an AI future where large, centralized models monopolize the market, widen inequality, and deplete our resources, versus one where small, diversified models are built from the bottom up, powered sustainably, and used to yield benefits across society. In this scenario, AI models will also be 

needing fewer data centers in the first place.

After all, this technology should be a means to an end, not the end itself. Slow and steady—and sustainable—wins the race.

, was made possible with support from the Pulitzer Center and is part of the 

, a cross-border collaborative investigation led by Brazilian news organization Agência Pública and the Centro Latinoamericano de Investigación Periodística (CLIP) with partners including Tech Policy Press.

For 66-year-old Ncediwe Vanqa, water comes one trip at a time. She walks from her less than 200 square foot dwelling in Barcelona, an informal settlement in Gugulethu, Cape Town, to a communal tap as many as seven times a day. She uses a communal pit latrine and, in emergencies, a bucket. To save electricity, she bakes bread using a paraffin heater and uses its heat to warm her home.

Yet across the freeway, less than a mile and a half away, the city of Cape Town has opened the planning door to infrastructure operating at an entirely different scale. In July, the City's Municipal Planning Tribunal 

 a planning application by King Air Industria, a multi-billion-rand industrial and logistics park, for two proposed data centers that planning documents identify as intended for Equinix, a US multinational.

Equinix, one of the world’s largest data-center operators and a key infrastructure provider for hyperscalers such as Amazon, Microsoft and Google, disputes that characterization.

Its expansion in Cape Town is the focus of this story, part of 

, a cross-border investigation examining how global technology companies are influencing the rules, policies and public infrastructure.

it “has not submitted any planning or development applications for a data center in Cape Town” and that “there are currently no development applications under consideration for the site.”

But the city’s own account shows that Equinix met officials at least three times to discuss its data-center plans. Dozens of planning documents also link the King Air land-use application to Equinix’s proposed development of the site.

At the center of the King Air dispute is the question of how much water and electricity the proposed Equinix hyperscale data centers will require. Hyperscale data centers are notorious worldwide for consuming inordinate amounts of water and/or electricity to cool the machines.

In its appeal against the building of one of these in Cape Town, the Legal Resources Center (LRC), a public-interest law organization, estimated that, if conventional evaporative cooling were used, the two data centers could consume about 1.2 billion gallons of water a year. That is almost 3 million gallons a day, roughly equivalent to the daily water use of 68,500 Cape Town residents, based on the city's average consumption 

. For a city that narrowly avoided “Day Zero” during the 2018 drought, the potential demand is significant.

Equinix did not respond to questions relating to the proposed data center resources. However, in its papers, King Air told the Tribunal that the final amounts of resource usage would be made available in the site development planning phase.

In the US, data center expansion is facing growing community resistance. According to 

, an independent nonprofit research and policy institute in the US, global data center energy consumption has grown 12% in five years and could double by 2030. Water use is also rising. The organization says consumption by Meta, Apple, Microsoft and Google has increased 60% since 2020. As data centers expand into communities, the organization argues that where they are built has become an increasingly political decision, with consequences for local economies, health, water and energy supplies.

According to the documents before the Municipal Planning Tribunal, the proposed system could need about 157 megawatts. “If operated continuously near this level,” throughout the year, the data centers could consume roughly 1.4 to 1.5 billion kilowatt hours of electricity annually, explained Dr. Chantelle van Staden, a senior lecturer at the Stellenbosch University’s Department of Electrical and Electronic Engineering. “This could amount to the annual electricity use of a quarter of a million average Cape Town households,” said Van Staden.

That lack of information troubled at least one member of the tribunal when it considered the land-use approval in July. “Most of us here today actually do not know what is required in respect of water and electricity demand for a facility that we are asked to approve, as no details thereof have been made available,” Municipal Planning Tribunal panel member Wally Johnstone said at a July 14 meeting. Johnstone was the only one of five panel members opposing the approval of land use for data centers.

The city disputes that the development was approved without adequate consideration and that its energy department had concluded that there would be sufficient supply. The tribunal papers also revealed that Equinix has already paid shared network charges for 144 megawatts of capacity amounting to R33 million (around $1.9 million). The city said some technical details would be assessed later.

In Barcelona, residents have to walk and use buckets to get water just a little more than a mile away from Equinix's data center. Photo: David Harrison/The Continent

Weeks before the tribunal approved the King Air application, Public Protector Kholeka Gcaleka released findings concerning municipal service delivery to informal settlements Langa and Khayelitsha, which are in close proximity to Equinix’s proposed data centers. Khayelitsha and Langa are both high-density, mostly poor townships less than 6.2 miles away from the proposed site. In June, after a four-year-long investigation, the 

 that Cape Town had failed to provide adequate basic services to residents of the two townships, including water, sanitation and electricity. The failures were serious enough to amount to “maladministration” and “prejudice”.

The city told the investigation that budget constraints, infrastructure backlogs and competing service-delivery demands had prevented it from fixing some of these problems. The public protector acknowledged the financial pressures but found they did not absolve the City of its constitutional obligations, with systemic service-delivery failures continuing despite some improvements.

For residents and Housing Assembly members such as Vanqa, the argument is about who gets scarce resources. The LRC told the tribunal that their objection originates from “a context of acute and overlapping resource constraints in the City of Cape Town where land, water and energy systems are already under significant pressure.”

Seven times a day: Ncediwe Vanqa uses buckets to access water in Barcelona, an informal settlement in Gugulethu, Cape Town. Photo: David Harrison/The Continent 

South Africa is not new to data centers. 

, an industry directory, lists about 64 operating data centers in South Africa and another 10 planned. But hyperscale facilities bring infrastructure demands of a different order, with electricity requirements that can run into hundreds of megawatts. The regulations have not kept pace. South Africa has no dedicated national data-center policy, while Cape Town has no separate land-use category for hyperscale facilities. The city said data centers are “typically grouped under commercial/industrial users," the city told 

. They are assessed under existing industrial or warehousing rules.

Johnstone called this “wholly inadequate” and added that it was “simplistic to consider a data center as a warehouse that only stores computer systems.” The LRC went further, describing the classification as “patently incorrect and possibly misleading.”

Speaking on how the application before the tribunal uses legislation that classifies data centers as warehouses, Grant Oosterwyk, a senior lecturer in information systems at the University of Cape Town who previously worked at a small-scale data center, said the current approach creates an artificial blind spot.

“This allows these facilities to avoid the rigorous, baseline industrial impact assessments that a facility of this scale would normally trigger under a modern planning framework,” Oosterwyk said.

“Treating a 100 megawatt facility with industrial-scale cooling, heavy diesel backup, and continuous heavy vehicle movement as a simple storage warehouse is a profound regulatory mismatch. The category has simply not caught up with the infrastructure being built inside it.”

What Cape Town won't disclose about its data center pipeline

The proposed Equinix development is also not the only hyperscale project being contemplated in Cape Town. Brett Herron, a member of the Western Cape Provincial Parliament, told 

that Cape Town is "courting" four hyperscale projects.

He identified Equinix's proposed King Air facilities, Teraco's planned 60 megawatt expansion and Cavaleros Group's announced 360 megawatt Cosmas Data Cities project. Herron calculates their potential combined demand at approximately 580 megawatts. Based on publicly available information, he calculates that if all the projects were built at the stated capacities, their combined demand would equal roughly 35% of Cape Town’s current peak electricity demand.

Deputy mayor Alderman Eddie Andrews said Cape Town has approved five data centers and has another two under assessment. But the city would not name the companies behind those projects, citing the Privacy of Personal Information Act (POPIA). However, it did not explain under which provisions of the act prevented it from disclosing electricity-load information or why project-level demand could not be released.

Aside from legislation, the city’s Lance Greyling, the City’s Director of Economic Development and Investment, confirmed that he met Equinix three times between July 2025 and June 2026, all at the company’s request.

At the first meeting in July 2025, Greyling said, Equinix told him of “their intention to invest in data centers in Cape Town.” Four months later, Greyling, mayoral committee member James Vos, city officials and Western Cape government representatives met the company after it had bought land at King Air Industria. Greyling said Equinix discussed “their intentions in relation to the way forward for the proposed business.” At a third meeting in June, the company provided a further update on its planned data-center investment.

Greyling said, “I have met Equinix before at their request but have at no stage held discussions on statutory approvals.” Yet there were no agendas or minutes kept. He said his role is to engage potential investors and promote Cape Town as a place for economic growth, provided they meet the necessary legal requirements.

In South Africa, meetings between officials and private investors are not routinely disclosed or regulated. There is no evidence Equinix influenced the statutory approval process, but the company met Cape Town’s investment-promotion office repeatedly as its data-center plans developed. With no minutes kept, there is no detailed public record of what was discussed.

interviewed two city councillors who said the implications of data center expansion had not been widely debated in the city council. Khalid Sayed, the ANC’s leader in the Western Cape, told 

that data centers, including their water and electricity demands, have barely featured in political forums. He said this may reflect both a lack of understanding and oversight failure.

Cape Town is not alone. eThekwini Municipality acknowledged that its planning scheme has no specific definition for hyperscale or AI data centers and said it was still considering how to regulate their infrastructure demands.

that its full council had authorized the city to enter into an agreement with the Korean South Power Consortium to explore a proposed AI data center on municipal land in the south of Durban. Council documents estimate the development could cost between $3 billion and $10 billion and state that the municipality would provide the land and infrastructure required for the project. The city said the project's water and electricity requirements have not yet been determined.

Marketing and communications director Mandla Nsele said the issues raised regarding data center demands "are the subject of internal discussions," meaning the municipality is unable to provide detailed data.

One of the council members against the development, Andre Beetge, said a council report regarding the project does not “adequately address the energy demand trajectory associated with AI data centers—projected to escalate dramatically over the next few years.”

The roundup is produced by Digital Policy Alert, an independent repository of policy changes affecting the digital economy. If you have feedback or questions, please contact 

The roundup serves as a guide for navigating global digital policy based on the work of the 

. To ensure trust, every finding links to the Digital Policy Alert entry with the official government source. The full Digital Policy Alert 

 is available for you to access, filter, and download. To stay updated, Digital Policy Alert also offers a customizable 

 that provides free updates on your areas of interest. Digital Policy Alert’s 

 further allow you to navigate, compare, and chat with the legal text of AI rules across the globe.

Drawing from Digital Policy Alert’s daily monitoring of developments in the G20 countries, it summarizes the highlights of August 2026 in four core areas of digital policy.

Content moderation, including the European Commission’s designation of ChatGPT as a very large online search engine and Reddit and Roblox as very large online platforms under the Digital Services Act, the French Constitutional Council’s decision finding unconstitutional provisions of a bill restricting social media access for children under 15, Brazil’s enforcement actions concerning Discord’s compliance with child-safety obligations, and United Kingdom Ofcom’s expansion of its investigations into Chat-Avenue and Teen-Chat.

AI regulation, including the entry into force of the EU AI Act's transparency rules and the AI Office's enforcement powers over general-purpose AI models, South Korea's adoption of AI Ethics Principles, Saudi Arabia's approval of the International Centre for Artificial Intelligence Research and Ethics, and Turkey's AI Action Plan.

Competition policy, including the European Commission's approval of six digital infrastructure and technology mergers, Germany's Federal Cartel Office closing its investigation into Apple's App Tracking Transparency Framework, and Turkey's Competition Authority approving the Warner Bros. Discovery-Paramount Skydance acquisition.

Data governance, including Brazil's National Data Protection Authority fine of R$ 153.7 million against ByteDance's TikTok, China's consultation on Regulations on Protection of Personal Information by Large-Scale Personal Information Processors, Japan's administrative guidance to KDDI following a breach affecting 12.2 million people, and Canada's Federal Court application against Google over search de-listing obligations.

 as a very large online search engine under the Digital Services Act (DSA), classifying it as a hybrid search service because it can respond to user queries by searching the web. At the same time, 

 were designated as very large online platforms, as they enable users to disseminate third-party content to the public. As a result, the three services must comply with the DSA's additional obligations by the end of January 2027. These include assessing and mitigating systemic risks posed by illegal content, protecting minors, safeguarding users' physical and mental well-being, upholding fundamental rights, ensuring the integrity of electoral processes, and maintaining public security.

 was introduced in the Austrian National Council to establish a blocking regime for prohibited online gambling offers, with fines of up to €1 million. A 

 regulating digital commercial influence was introduced to Belgium's Parliament, which would require hosting providers to operate illegal-content reporting mechanisms, publish transparency reports, prioritize notices from 

 obligations. Regarding enforcement in Hungary, the Competition Authority fined the e-commerce platform 

 HUF 225 million for failing to fulfill earlier commitments to support Hungarian businesses.

 Constitutional Council found the provisions of the 

 banning minors under 15 from accessing social media unconstitutional. The Council held that the ban was disproportionate to its objective of protecting minors and that the age-verification requirement lacked sufficient legal safeguards for the processing of personal data.

, the Communications Regulatory Authority's supplementary 

 on communication campaigns against gambling disorders entered into force. The guidelines establish rules for advertising across digital channels, including short-form social media videos and automated messages to registered players, and distinguish responsible gambling communications from prohibited promotional advertising.

 requiring the labeling of content generated or modified using large foundation AI models was introduced in the State Duma. If adopted, the Government would subsequently determine the format, content, and procedure for applying the labels.

Latest

Podcast

Recent

Our content delivered to your inbox

Thank you!

Around the globe