{"id":17391,"date":"2026-02-05T12:02:04","date_gmt":"2026-02-05T11:02:04","guid":{"rendered":"https:\/\/haimagazine.com\/uncategorized\/ais-green-tax-will-data-centers-face-special-taxation\/"},"modified":"2026-02-10T09:37:56","modified_gmt":"2026-02-10T08:37:56","slug":"ais-green-tax-will-data-centers-face-special-taxation","status":"publish","type":"post","link":"https:\/\/haimagazine.com\/en\/business-2\/ais-green-tax-will-data-centers-face-special-taxation\/","title":{"rendered":"\ud83d\udd12 AI&#8217;s green tax: Will data centers face special taxation?"},"content":{"rendered":"<p>A few years ago, the question &#8220;does AI use too much electricity?&#8221; was treated as a marginal note in drafts of ESG reports. Today the answer is obvious: it does\u2014and at a pace that threatens the stability of power grids. The question that regulators from Brussels to Washington are now asking is different: who will pay for it?<\/p><p>Estimates by international institutions suggest that a special tax on energy consumption by data centers could generate billions of dollars annually, correcting market distortions. According to experts, the current price of AI services is artificially low because it doesn&#8217;t account for externalities\u2014such as air pollution or water consumption\u2014that are ultimately borne by citizens rather than technology companies. For business, this is a wake-up call. If AI models become subject to targeted taxation, the entire economics of deployments, from chatbots to advanced predictive systems, could change drastically.<\/p><h4 class=\"wp-block-heading\">The scale dilemma<\/h4><p>In this dispute, it&#8217;s not about whether to protect the environment, but how to design fiscal mechanisms so as not to throw the baby out with the bathwater. This is the core of the debate that took place during Generative Revolution Day \u2014 <br\/><mark style=\"background-color:#82D65E\" class=\"has-inline-color has-base-color\"><a href=\"https:\/\/generativerevolution.ai\/should-AI-pay-higher-taxes\/\" target=\"_blank\" rel=\"noopener\">Should AI companies pay higher taxes because of their impact on the environment?<\/a><\/mark> \u2014 which featured Paz Pe\u00f1a, a socio-technical justice expert, and Luis Bitencourt-Emilio, a veteran of the tech industry with experience at Microsoft and Reddit.<\/p><p><em>The rest of the article is below the video<\/em><\/p><figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"Should AI companies pay higher taxes because of their impact on the environment?\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/ixDaMq1Wyvo?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure><p><\/p><p>For Pe\u00f1a, data centers are not ordinary factories, and the problem goes beyond just electricity.<\/p><p>&#8220;AI data centers are highly controversial. Their intensive computing power requires increasing amounts of energy and, as a result, more freshwater to cool servers and infrastructure,&#8221; the expert notes. She points to the United States, which hosts over 50% of the world\u2019s data centers, as a testing ground where these effects are already visible.<\/p><p>In her view, rapid digital expansion poses a real threat to the energy transition, prolonging the life of fossil fuels.<\/p><p>&#8220;In fact, we\u2019re seeing the opposite of a transition. The enormous energy demand from data centers is an ideal opportunity to develop new fossil-fuel projects,&#8221; says Pe\u00f1a. Citing the Institute for Energy Economics and Financial Analysis, Pe\u00f1a points out that the digital boom is driving the construction of gas-fired power plants and pipelines, which clearly hinders the energy transition.<\/p><p>The key argument for introducing a targeted tax (ranging from $0.032 to $0.052 per kilowatt-hour) is the mechanism that Pe\u00f1a describes as socializing costs while privatizing profits. In practice, this means that while tech giants reap the profits from deploying AI, the bill for the side effects\u2014such as higher electricity prices for households, consumption of local water resources, or polluted air\u2014is paid by ordinary citizens.<\/p><p>&#8220;This could generate up to $18 billion a year,&#8221; the expert estimates, also emphasizing the need for international coordination: &#8220;The tax should be global to avoid placing data centers in locations with weaker regulations.&#8221;<\/p><p>Pe\u00f1a also points out the balance of gains and losses for local communities: <\/p><p>&#8220;Given the environmental damage, the negligible impact on job creation and the pressure on power grids, which can increase prices for households and reduce demand for other goods, the net benefits of these special tax regimes are at best uncertain,&#8221; she adds.<\/p><h4 class=\"wp-block-heading\">Measurement trap<\/h4><p>Luis Bitencourt-Emilio takes a markedly different approach. In his view, taxing a specific technology is a logical fallacy that leads to the so-called &#8220;measurement trap.&#8221; AI is a &#8220;general-purpose technology,&#8221; like electricity, and pervades almost every sector of the economy.<\/p><p>&#8220;What would you tax? Model training, inference, AI embedded in applications, chips, cars, airplanes, hospitals, phones?,&#8221; asks Bitencourt-Emilio, exposing the legislative challenges. &#8220;Defining boundaries that will be indisputable across industries is prone to errors and manipulation. Methods differ on what to include, how to allocate shared infrastructure, and how to normalize the data: per token, per query, or per model?&#8221;<\/p><p>Instead of creating complicated definitions, the expert proposes a simple market mechanism: taxing CO2 emissions at the source. <\/p><p>&#8220;There is no black-and-white answer. But if you care about the environment, the right move is to price pollution, not innovation,&#8221; he argues. Bitencourt-Emilio cites hard economic data: &#8220;The Commission on Carbon Prices recommends a pathway of $50 to $100 per ton of carbon dioxide by 2030 to align with climate goals.&#8221;<\/p><p>So his vision boils down to a simple solution: &#8220;If AI scales, it naturally pays more when it drives demand. If it runs on cleaner energy, it pays less. You get the right incentives without stifling a general-purpose technology.&#8221;<\/p><h4 class=\"wp-block-heading\">Theory meets reality<\/h4><p>To understand the real impact of these proposals on business, let&#8217;s analyze them with an example. Let&#8217;s assume there&#8217;s a hypothetical company &#8220;DataCenter X&#8221; that consumes 100 GWh of energy annually, half of which sourced from fossil fuels.<\/p><p>In the scenario proposed by Pe\u00f1a (a fixed earmarked tax), the company is suddenly burdened with an additional cost of about $5 million per year. This charge is difficult to absorb in the short term from its own funds, which forces the company to raise the prices of its services.<\/p><p>The Bitencourt-Emilio market scenario\u2014a high carbon tax that exempts clean energy\u2014seems more rational, but when confronted with infrastructural reality, it reveals its shortcomings. Theoretically, DataCenter X should sign PPAs (Power Purchase Agreements) for green energy to reduce the tax to zero. However, in practice the company falls into an &#8220;availability trap,&#8221; which seems to confirm Pe\u00f1a&#8217;s concerns.<\/p><p>What\u2019s the issue? In key technology hubs, such as Virginia or Frankfurt, green energy is a scarce commodity. A tech giant buying up what little is left would drastically drive up electricity prices for local residents. Therefore, the company has to finance new sources, but here it runs up against timelines and bureaucracy. Connecting a new wind farm to the transmission grid in the US and EU takes currently 3-5 years. This creates the most challenging situation for businesses: during the transition period, the company must finance investments in renewables while still paying punitive taxes for using dirty energy.<\/p><h4 class=\"wp-block-heading\">The cheap days are over&#8230;?<\/h4><p>Regardless of whether regulators opt for a fiscal or market model, the days when the cost of a query to an AI model was negligible seem to be coming to an end. The environmental cost will inevitably be included in the token price. Companies that ignore this fact risk not only financial but also reputational losses. As Paz Pe\u00f1a notes, the current system is unfair:<\/p><p>&#8220;Countries lose tax revenue while simultaneously financing infrastructure for companies that increase global emissions.&#8221;<\/p><p>The only way out seems to be radical optimization. Bitencourt-Emilio cites an example from history: <\/p><p>&#8220;We solved the pollution problem by targeting sources rather than banning technologies. We didn&#8217;t tax cars or flights per se. American cars are now 99% cleaner in terms of key pollutants than models from 1970.&#8221;<\/p><p>In his view, the same awaits AI, provided that regulations demand engineering ingenuity instead of imposing a fiscal clampdown.<\/p><p>For IT managers, this means a paradigm shift. The energy efficiency of code is becoming a key performance indicator (KPI). Implementing a precise &#8220;AI Carbon Accounting&#8221; (monitoring the carbon footprint per project) will be essential. Moreover, the technology strategy will need to account for energy geography, and ambitions will give way to specialization. In a world of expensive energy, using premium-class models for simple classification tasks will be seen as business recklessness, like using a cannon to kill a fly. Therefore, the future belongs to smaller, specialized models that consume a fraction of the energy of their larger cousins. Such architectural optimization not only lowers the cloud bill but, above all, is the most effective shield against impending environmental taxes. And it can be implemented right away, without any investment in green infrastructure.<\/p>","protected":false},"excerpt":{"rendered":"<p>The issue of AI\u2019s impact on the climate has ceased to be the domain of activists and has made it onto CFOs&#8217; agendas. Is AI another industrial revolution that requires freedom, or an ecological parasite that should be taxed at the source? Two visions of the future are clashing in debates, with the profitability of the digital economy at stake.<\/p>\n","protected":false},"author":354,"featured_media":17345,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"rank_math_lock_modified_date":false,"footnotes":""},"categories":[888],"tags":[],"popular":[],"difficulty-level":[],"ppma_author":[776],"class_list":["post-17391","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-business-2"],"acf":[],"authors":[{"term_id":776,"user_id":354,"is_guest":0,"slug":"redakcja","display_name":"Redakcja","avatar_url":{"url":"https:\/\/haimagazine.com\/wp-content\/uploads\/2025\/07\/Zrzut-ekranu-2025-07-10-o-16.00.36.png","url2x":"https:\/\/haimagazine.com\/wp-content\/uploads\/2025\/07\/Zrzut-ekranu-2025-07-10-o-16.00.36.png"},"first_name":"","last_name":"","user_url":"","job_title":"","description":""}],"_links":{"self":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/17391","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/users\/354"}],"replies":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/comments?post=17391"}],"version-history":[{"count":1,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/17391\/revisions"}],"predecessor-version":[{"id":17392,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/17391\/revisions\/17392"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/media\/17345"}],"wp:attachment":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/media?parent=17391"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/categories?post=17391"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/tags?post=17391"},{"taxonomy":"popular","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/popular?post=17391"},{"taxonomy":"difficulty-level","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/difficulty-level?post=17391"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/ppma_author?post=17391"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}