{"id":45,"date":"2026-02-19T16:25:31","date_gmt":"2026-02-19T16:25:31","guid":{"rendered":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/?p=45"},"modified":"2026-09-14T12:49:47","modified_gmt":"2026-09-14T11:49:47","slug":"are-tres-a-bad-idea","status":"publish","type":"post","link":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/are-tres-a-bad-idea\/","title":{"rendered":"Are TREs a bad idea?"},"content":{"rendered":"<p style=\"font-weight: 400\">Trusted Research Environments \u2014 or\u00a0<em>TREs<\/em>\u00a0\u2014 are often pitched as the future of safe data access. Governments love them, big research programmes promote them, and policy papers talk about them like they\u2019re the silver bullet for patient privacy.<\/p>\n<p style=\"font-weight: 400\">But when you look past the branding, TREs come with some serious drawbacks. In fact, in many everyday research settings, they can slow innovation, complicate collaboration, and introduce more barriers than benefits.<\/p>\n<p style=\"font-weight: 400\">So let\u2019s break down the problems \u2014 plainly, practically, and with evidence.<\/p>\n<h2><strong>1. TREs make research slower and more bureaucratic<\/strong><\/h2>\n<p style=\"font-weight: 400\">One of the biggest issues with TREs is that they\u00a0<strong>multiply the amount of admin researchers need to do<\/strong>, especially when a project requires data from more than one environment. Rather than pulling approved data into a single workspace, researchers often need:<\/p>\n<ul style=\"font-weight: 400\">\n<li>separate access applications<\/li>\n<li>separate technical checks<\/li>\n<li>separate governance approvals<\/li>\n<li>separate log\u2011ins<\/li>\n<li>separate restrictions on what tools they can use<\/li>\n<\/ul>\n<p style=\"font-weight: 400\">This can add\u00a0<strong>months<\/strong>\u00a0to a project timeline.\u00a0<a href=\"https:\/\/www.phgfoundation.org\/wp-content\/uploads\/2024\/11\/The-federation-of-trusted-research-environments-for-genomics-and-health-.pdf\">[phgfoundation.org]<\/a><\/p>\n<p style=\"font-weight: 400\">If your research involves multimodal data (e.g., imaging + genomics + clinical records) stored across different TREs, the situation becomes even worse: you might be stuck stitching together results in awkward formats because the systems don\u2019t talk to each other.\u00a0<a href=\"https:\/\/www.phgfoundation.org\/wp-content\/uploads\/2024\/11\/The-federation-of-trusted-research-environments-for-genomics-and-health-.pdf\">[phgfoundation.org]<\/a><\/p>\n<p style=\"font-weight: 400\">In fast\u2011moving fields \u2014 like cancer research, infectious disease, or AI model development \u2014 TRE-related delays aren\u2019t just annoying. They can make a study unviable.<\/p>\n<h2><strong>2. They\u2019re often \u201ctoo closed\u201d to be genuinely useful<\/strong><\/h2>\n<p style=\"font-weight: 400\">To make data \u201csafe,\u201d TREs lock it down. But sometimes they lock it down\u00a0<strong>too much<\/strong>.<\/p>\n<p style=\"font-weight: 400\">TREs typically restrict:<\/p>\n<ul style=\"font-weight: 400\">\n<li>moving data out<\/li>\n<li>bringing external tools in<\/li>\n<li>running custom code<\/li>\n<li>exporting large outputs<\/li>\n<li>collaborative work across institutions<\/li>\n<\/ul>\n<p style=\"font-weight: 400\">This means researchers may not be able to:<\/p>\n<ul style=\"font-weight: 400\">\n<li>use modern analytics tools<\/li>\n<li>reproduce previous analyses<\/li>\n<li>validate methods with external teams<\/li>\n<li>do cross\u2011centre quality assurance<\/li>\n<\/ul>\n<p style=\"font-weight: 400\">Several policy reviews note that TRE \u201cclosedness\u201d\u00a0<strong>directly inhibits analysis across environments<\/strong>\u00a0and makes combining results slow and expensive.\u00a0<a href=\"https:\/\/www.phgfoundation.org\/wp-content\/uploads\/2024\/11\/The-federation-of-trusted-research-environments-for-genomics-and-health-.pdf\">[phgfoundation.org]<\/a><\/p>\n<p style=\"font-weight: 400\">Ironically, TREs can hurt research quality while aiming to protect data quality.<\/p>\n<h2><strong> TREs don\u2019t actually \u201cbuild trust\u201d \u2014 they remove the need for it<\/strong><\/h2>\n<p style=\"font-weight: 400\">Here\u2019s something surprising: the whole \u201ctrusted\u201d part of Trusted Research Environments doesn\u2019t really hold up under ethical scrutiny.<\/p>\n<p style=\"font-weight: 400\">A Journal of Medical Ethics analysis argues that TREs\u00a0<strong>don\u2019t build trust at all<\/strong>\u00a0\u2014 they eliminate the vulnerability that trust requires, because everything is locked down, audited, and restricted.\u00a0<a href=\"https:\/\/jme.bmj.com\/content\/49\/10\/670\">[jme.bmj.com]<\/a><\/p>\n<p style=\"font-weight: 400\">Trust comes from transparency, professionalism, and accountability \u2014 not from designing a system where trust becomes irrelevant because no one has meaningful autonomy. TREs solve a technical risk but create a social one: the public is asked to \u201ctrust the system\u201d rather than the people running it.<\/p>\n<p style=\"font-weight: 400\">This makes the name itself somewhat misleading.<\/p>\n<h2 style=\"font-weight: 400\"><strong>TREs can block innovation at scale<\/strong><\/h2>\n<p style=\"font-weight: 400\">TREs were never designed for huge, modern datasets like:<\/p>\n<ul style=\"font-weight: 400\">\n<li>whole-genome sequences<\/li>\n<li>national imaging repositories<\/li>\n<li>drug\u2011discovery screens<\/li>\n<li>population\u2011scale multi\u2011omics<\/li>\n<\/ul>\n<p style=\"font-weight: 400\">According to a PHG Foundation briefing, TREs cause\u00a0<strong>major problems for large\u2011scale and multimodal datasets<\/strong>, especially in genomics where data volumes hit petabytes. Even just storing these datasets centrally creates\u00a0<strong>massive energy use, financial cost, and sustainability concerns<\/strong>.\u00a0<a href=\"https:\/\/www.phgfoundation.org\/wp-content\/uploads\/2024\/11\/The-federation-of-trusted-research-environments-for-genomics-and-health-.pdf\">[phgfoundation.org]<\/a><\/p>\n<p style=\"font-weight: 400\">As big datasets get bigger, TREs become a bottleneck \u2014 not a solution.<\/p>\n<h2 style=\"font-weight: 400\"><strong>They don\u2019t play well together (yet)<\/strong><\/h2>\n<p style=\"font-weight: 400\">A huge issue is that TREs across organisations or nations\u00a0<strong>aren\u2019t federated<\/strong>. This means:<\/p>\n<ul style=\"font-weight: 400\">\n<li>No unified login<\/li>\n<li>No shared technical standards<\/li>\n<li>No standardised governance<\/li>\n<li>No consistent analytical tooling<\/li>\n<\/ul>\n<p style=\"font-weight: 400\">So researchers must effectively operate in\u00a0<strong>multiple incompatible mini\u2011ecosystems<\/strong>, repeating work each time.\u00a0<a href=\"https:\/\/www.phgfoundation.org\/wp-content\/uploads\/2024\/11\/The-federation-of-trusted-research-environments-for-genomics-and-health-.pdf\">[phgfoundation.org]<\/a><\/p>\n<p style=\"font-weight: 400\">Several expert groups argue that TREs are\u00a0<em>not future\u2011proof<\/em>\u00a0unless they become interoperable \u2014 and currently, most are not.\u00a0<a href=\"https:\/\/www.phgfoundation.org\/wp-content\/uploads\/2024\/11\/The-federation-of-trusted-research-environments-for-genomics-and-health-.pdf\">[phgfoundation.org]<\/a><\/p>\n<h2 style=\"font-weight: 400\"><strong>TREs are expensive, resource\u2011intensive, and hard to scale<\/strong><\/h2>\n<p style=\"font-weight: 400\">Setting up a TRE requires:<\/p>\n<ul style=\"font-weight: 400\">\n<li>secure hosting<\/li>\n<li>specialised software<\/li>\n<li>strict monitoring<\/li>\n<li>trained administrators<\/li>\n<li>ethics and access management teams<\/li>\n<li>continual auditing and compliance checks<\/li>\n<\/ul>\n<p style=\"font-weight: 400\">This makes TREs disproportionately expensive for smaller institutions, charities, NHS services, or academic groups.<\/p>\n<p style=\"font-weight: 400\">Even at the national level, building multiple TREs has created fragmented, duplicated systems rather than a single, efficient platform. Many groups (including ARDC in Australia) acknowledge that TREs require heavy coordination just to stay functional.\u00a0<a href=\"https:\/\/ardc.edu.au\/program\/trusted-research-environments\/\">[ardc.edu.au]<\/a><\/p>\n<h2 style=\"font-weight: 400\"><strong>So\u2026 are TREs always bad?<\/strong><\/h2>\n<p style=\"font-weight: 400\">Not at all.<\/p>\n<p style=\"font-weight: 400\">TREs are a good fit when:<\/p>\n<ul style=\"font-weight: 400\">\n<li>handling\u00a0<em>highly sensitive<\/em>\u00a0data<\/li>\n<li>dealing with\u00a0<em>small, well\u2011defined<\/em>\u00a0research teams<\/li>\n<li>working within\u00a0<em>a single institution\u2019s dataset<\/em><\/li>\n<li>clear audit trails and tight control are essential<\/li>\n<\/ul>\n<p style=\"font-weight: 400\">But for most everyday research, especially multi\u2011centre or data\u2011intensive studies, the downsides far outweigh the supposed benefits.<\/p>\n<p style=\"font-weight: 400\">TREs can be:<\/p>\n<ul style=\"font-weight: 400\">\n<li>too slow<\/li>\n<li>too fragmented<\/li>\n<li>too heavily locked down<\/li>\n<li>too technically limiting<\/li>\n<li>too bureaucratic<\/li>\n<\/ul>\n<p style=\"font-weight: 400\">And although they protect data, they can choke the research that data is supposed to support.<\/p>\n<h2 style=\"font-weight: 400\"><strong>Final thought<\/strong><\/h2>\n<p style=\"font-weight: 400\">TREs aren\u2019t a magic answer \u2014 and in most circumstances, they\u2019re not even a good one. They solve important privacy concerns, but at the cost of agility, collaboration, innovation and, ironically, genuine trust.<\/p>\n<p style=\"font-weight: 400\">If we want research systems that are safe\u00a0<em>and<\/em>\u00a0scientifically productive, we need to think beyond simply locking data away \u2014 and build environments that support both privacy and progress.<\/p>\n<p style=\"font-weight: 400\">\n","protected":false},"excerpt":{"rendered":"Trusted Research Environments \u2014 or\u00a0TREs\u00a0\u2014 are often pitched as the future of safe data access. Governments love them, big research [&hellip;]","protected":false},"author":7176,"featured_media":47,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[5],"tags":[],"class_list":["post-45","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-accessing-data"],"meta_box":[],"_links":{"self":[{"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/posts\/45","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/users\/7176"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/comments?post=45"}],"version-history":[{"count":1,"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/posts\/45\/revisions"}],"predecessor-version":[{"id":46,"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/posts\/45\/revisions\/46"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/media\/47"}],"wp:attachment":[{"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/media?parent=45"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/categories?post=45"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blogs.cardiff.ac.uk\/bioresource-data-accelerator\/wp-json\/wp\/v2\/tags?post=45"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}