[{"data":1,"prerenderedAt":1038},["ShallowReactive",2],{"article-/newsroom/a-strategy-for-ai-in-education":3,"related-/newsroom/a-strategy-for-ai-in-education":208},{"id":4,"title":5,"author":6,"body":7,"date":178,"description":179,"documentId":180,"experienceName":180,"experienceUrl":180,"extension":181,"faqs":182,"hideContactCta":198,"image":199,"lastModified":180,"meta":200,"navigation":198,"path":201,"report":202,"seo":203,"seoDescription":180,"seoTitle":180,"stem":204,"subtitle":205,"tags":180,"topic":206,"__hash__":207},"newsroom/newsroom/a-strategy-for-ai-in-education.md","Introducing: A Strategy for AI & Education","Angel Maldonado",{"type":8,"value":9,"toc":168},"minimark",[10,19,22,27,30,33,37,40,43,46,50,53,56,63,66,70,81,85,88,91,100,105,109,112],[11,12,13,14,18],"p",{},"We've just published something we've been building for a while: a full strategic blueprint for how educational institutions can approach artificial intelligence deliberately, instead of reactively. It's called ",[15,16,17],"em",{},"A Strategy for AI & Education",", and this piece is the short version: what's in it, why we built it this way, and who it's for.",[11,20,21],{},"The short answer to \"why does this exist\" is that most institutions are being asked to have an AI policy before anyone has given them a coherent way to think about what AI actually does to a learner's mind. So they either lock it out entirely, or let it in with no more structure than a chatbot window and a hope for the best. Neither is a strategy. Both are ways of avoiding the actual question.",[23,24,26],"h2",{"id":25},"it-starts-with-a-genuine-tension-not-a-sales-pitch","It starts with a genuine tension, not a sales pitch",[11,28,29],{},"The blueprint opens where we think any honest conversation about AI in education has to start, with the real possibility that AI in the classroom quietly optimizes for the wrong thing. Not because anyone intends harm, but because efficiency and engagement are easy to measure and curiosity isn't. A system built to reduce friction will reduce friction, including the productive kind, the struggle that's actually how understanding gets built. Left unexamined, that's not education. It's a faster way to arrive at the same shallow place.",[11,31,32],{},"That's the challenge. The response we lay out isn't \"reject AI.\" It's AI repositioned as a tool that operates inside an institution's own agency, governance, and control, rather than the institution operating inside whatever a vendor decided to ship.",[23,34,36],{"id":35},"then-it-gets-specific-with-actual-evidence","Then it gets specific, with actual evidence",[11,38,39],{},"This is the part we're most glad we included, and most careful about how we framed it. The blueprint devotes real space to the MIT Media Lab's 2025 study on \"cognitive debt,\" the EEG research showing that brain connectivity scaled down with how much AI support a writer used, and that 78% of habitual AI users couldn't quote a line from an essay they'd personally written minutes before.",[11,41,42],{},"We didn't want to cherry-pick the scary number and move on. So we went further into the same study and pulled out its most useful finding: when participants who'd written unaided were finally given AI access, their neural engagement spiked higher than anything the habitual-AI group ever produced. Timing, not exposure, was what moved the result. The paper's own conclusion is that AI should be introduced after self-driven effort, not instead of it, and that's the single idea the rest of the blueprint is built to operationalize.",[11,44,45],{},"From there we lay out three reframes we think matter more than the usual \"AI good, AI bad\" framing: scaffolding instead of offloading, neuroplastic enhancement instead of atrophy, and active metacognition instead of passive consumption. Each one turns out to hinge on something a piece of software can either be built to do or not, which is exactly why governance, not sentiment, is the lever that actually moves outcomes.",[23,47,49],{"id":48},"what-governance-looks-like-when-its-not-just-a-word","What governance looks like when it's not just a word",[11,51,52],{},"The blueprint gets concrete fast after that. We map four capability lines any serious AI-in-education engine needs: agentic governance over how and when AI activates, pattern analysis of real learning evidence, provenance and sources on every answer, and generative support for the reports and feedback that actually eat a teacher's week. Each one answers a specific finding from the research, not just a generic wish list.",[11,54,55],{},"Underneath that sits the actual architecture: Empathy.AI Spaces running on a Local AI Workstation, with a reference design any institution's technical team can evaluate on its own terms: knowledge storage, retrieval, interchangeable reasoning engines, and a capacity model built for how a school, not a single user, actually works. We included the hardware and the economics deliberately, because \"own your infrastructure\" is a much smaller idea if nobody shows their work on what that costs and what it actually buys.",[11,57,58],{},[59,60],"img",{"alt":61,"src":62},"Black Empathy.AI workstation on a table in a classroom","/media/newsroom/article14_ai-education-workstation.webp",[11,64,65],{},"A Local AI Workstation an institution can own: the school holds the machine, and the machine holds the school's knowledge.",[23,67,69],{"id":68},"we-didnt-want-to-argue-this-in-the-abstract","We didn't want to argue this in the abstract",[11,71,72,73,76,77,80],{},"So the blueprint includes three live examples, at different scales. ",[15,74,75],{},"Molly Vs The Machines"," shows the pattern in miniature: grounded retrieval plus a genuinely agentic document-generation feature, running today. ",[15,78,79],{},"Project Gutenberg AI"," shows the same architecture holding up across a genuinely large, heterogeneous knowledge base. And Putney High School, a GDST school running the full setup on its own campus, shows what age-staged governance actually looks like in a real institution: closed, source-cited first contact for younger students, and supervised, student-built development further up the school, all under policy the school wrote itself rather than terms it clicked through.",[23,82,84],{"id":83},"what-it-isnt","What it isn't",[11,86,87],{},"It isn't a sales deck dressed up as research, and it isn't a claim that owning hardware automatically fixes cognitive debt. We say so directly in the document. Local AI, left ungoverned, carries the same risks as anyone else's chatbot. What ownership actually buys an institution is the ability to enforce the sequence the evidence points to, instead of inheriting whichever sequence a vendor found most profitable. That's a meaningfully smaller claim than \"AI ownership solves education,\" and we think it's the honest one.",[11,89,90],{},"It also isn't finished, in the sense that matters. The last page of the document says so explicitly: this is a blueprint, not a prescription. Every institution's data, platforms, and pedagogical priorities are different, and the right pace of adoption should be built around them, not around us.",[11,92,93,94,99],{},"If any of that sounds like the conversation your school or institution has been putting off, the full document is where we've put the actual thinking. We'd like to ",[95,96,98],"a",{"href":97},"/contact/","have that conversation with you",".",[101,102],"newsroom-contact-cta",{"className":103},[104],"my-10",[23,106,108],{"id":107},"references","References",[11,110,111],{},"This article and the strategy document it summarizes were drafted with the assistance of AI, then reviewed and directed by a human in the loop. Claims and figures throughout are drawn from the following sources:",[113,114,115,119,124,130,136,142,148,154],"ul",{},[116,117,118],"li",{},"Kosmyna, N., Hauptmann, E., Yuan, Y.T., Situ, J., Liao, X.-H., Beresnitzky, A.V., Braunstein, I., Maes, P. — \"Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task,\" MIT Media Lab, preprint, June 2025",[116,120,121,123],{},[15,122,17],{}," (Empathy.AI), the full strategic blueprint this article summarizes",[116,125,126,129],{},[15,127,128],{},"AI in Education: An Empathetic Vision for Creativity, Art, and Liberation"," (Empathy.AI)",[116,131,132,135],{},[15,133,134],{},"AI Governance in the Perspective of Empathy AI"," (Empathy.AI, July 2026)",[116,137,138,141],{},[15,139,140],{},"Edge AI for Education"," (Empathy.AI), workshop programme documentation",[116,143,144,147],{},[15,145,146],{},"Local AI Workstation, v3"," (Empathy.AI), technical dossier",[116,149,150,153],{},[15,151,152],{},"About Empathy.AI Spaces",", product overview and deployment options",[116,155,156,157,162,163,167],{},"Case study material for ",[95,158,75],{"href":159,"rel":160},"https://mollyvsthemachines.com/",[161],"nofollow",", ",[95,164,79],{"href":165,"rel":166},"https://projectgutenberg.empathy.ai/",[161],", and Putney High School (Girls' Day School Trust)",{"title":169,"searchDepth":170,"depth":170,"links":171},"",2,[172,173,174,175,176,177],{"id":25,"depth":170,"text":26},{"id":35,"depth":170,"text":36},{"id":48,"depth":170,"text":49},{"id":68,"depth":170,"text":69},{"id":83,"depth":170,"text":84},{"id":107,"depth":170,"text":108},"2026-09-29","A blueprint for institutions that want to meet AI head-on, without asking students to trade their thinking for it.",null,"md",[183,186,189,192,195],{"question":184,"answer":185},"Why does a strategy for AI in education need to exist?","Most institutions are asked to have an AI policy before anyone has given them a coherent way to think about what AI does to a learner’s mind. Locking it out, or letting it in as a chatbot window, is not a strategy. Both avoid the actual question.",{"question":187,"answer":188},"What did the MIT Media Lab 2025 study find about cognitive debt?","EEG research showed that brain connectivity scaled down with how much AI support a writer used, and that 78% of habitual AI users could not quote a line from an essay they had written minutes before. When participants who had written unaided were then given AI, their neural engagement spiked higher than the habitual-AI group. Timing, not exposure, moved the result. The paper’s conclusion is that AI should be introduced after self-driven effort, not instead of it.",{"question":190,"answer":191},"What capability lines does the blueprint set for an education AI engine?","Four lines: agentic governance over how and when AI activates, pattern analysis of real learning evidence, provenance and sources on every answer, and generative support for the reports and feedback that take a teacher’s week. Each one answers a finding from the research.",{"question":193,"answer":194},"Which live examples does the strategy include?","Molly Vs The Machines shows grounded retrieval plus agentic document generation. Project Gutenberg AI shows the same architecture across a large, heterogeneous knowledge base. Putney High School, a GDST school, runs the full setup on its own campus, with age-staged governance under policy the school wrote.",{"question":196,"answer":197},"What does owning a local AI workstation not claim to fix?","Owning hardware does not automatically fix cognitive debt. Local AI, left ungoverned, carries the same risks as any other chatbot. Ownership buys the ability to enforce the sequence the evidence points to, instead of inheriting a sequence a vendor found most profitable.",true,"/media/newsroom/article14_ai-education.webp",{},"/newsroom/a-strategy-for-ai-in-education","strategy-for-ai-and-education",{"title":5,"description":179},"newsroom/a-strategy-for-ai-in-education","A blueprint for institutions that want to meet AI head-on, without asking students to trade their thinking for it","Education","sRj1R9BLZxZwQhGl4hXdfReQEH7gxh0AyADSYlebj9c",[209,400,751],{"id":210,"title":211,"author":6,"body":212,"date":370,"description":371,"documentId":180,"experienceName":180,"experienceUrl":180,"extension":181,"faqs":372,"hideContactCta":198,"image":390,"lastModified":180,"meta":391,"navigation":198,"path":392,"report":393,"seo":394,"seoDescription":180,"seoTitle":395,"stem":396,"subtitle":397,"tags":180,"topic":398,"__hash__":399},"newsroom/newsroom/a-strategy-for-ai-in-research.md","Introducing: A Strategy for AI in Research",{"type":8,"value":213,"toc":362},[214,220,223,227,230,233,240,244,247,250,253,257,260,263,269,273,276,279,282,293,297,300,303,308,311,313,316],[11,215,216,217,18],{},"We've just published a strategic blueprint for how research institutions can approach artificial intelligence without trading authorship for speed. It's called ",[15,218,219],{},"A Strategy for AI in Research",[11,221,222],{},"AI can help researchers find evidence, explore methods and question their conclusions. Its value depends on whether that assistance makes the research stronger, not simply faster.",[23,224,226],{"id":225},"why-does-this-strategy-exist","Why does this strategy exist?",[11,228,229],{},"The short answer is that most research teams are being asked to have an AI policy before anyone has given them a coherent way to think about what AI does to a claim. So they either lock it out, or let it in with no more structure than a chatbot window and a hope that a fluent answer is a justified one. Neither is a strategy. Both avoid the actual question.",[11,231,232],{},"The question the blueprint organises around is simple: what supports this claim? A convincing answer is not enough. Researchers should be able to see what supports a statement, where the evidence is missing, and which decisions remain theirs.",[11,234,235,236,239],{},"That is not an abstract worry. Messeri and Crockett's 2024 ",[15,237,238],{},"Nature"," Perspective describes how AI can encourage illusions of understanding and scientific monocultures, in which particular methods and viewpoints displace alternatives. The danger is not that AI is present in the lab. It is that a fluent result becomes a substitute for the intellectual work needed to justify it.",[23,241,243],{"id":242},"what-does-authored-knowledge-at-the-centre-mean","What does authored knowledge at the centre mean?",[11,245,246],{},"This strategy puts authored knowledge at the centre. In a knowledge-centric research workflow, approved protocols, literature, datasets and research records provide the evidence context. The system should retrieve relevant material, expose authors and versions, and distinguish what the sources say from what the model infers.",[11,248,249],{},"When evidence runs out, the preferred response is 'not found in this collection', followed by a way to inspect the search or extend its scope. Conjecture remains legitimate. It should be labelled as exploration, and verified before anyone accepts it as a finding.",[11,251,252],{},"We contrast that with an LLM-centric workflow that has no access to the project archive. Asked which exclusion rule was approved for a study, it can invent a plausible rule and, on a follow-up, invent an amendment that never existed. The researcher still has to check the source. The left-hand behaviour is an acceptance requirement. The right-hand answers illustrate a failure mode.",[23,254,256],{"id":255},"does-local-ai-make-an-answer-correct","Does local AI make an answer correct?",[11,258,259],{},"No. Local AI gives institutions greater control over their research environment. Keeping documents, prompts, embeddings, retrieved passages, outputs and inference within an institutional boundary can reduce exposure. It does not make an answer correct.",[11,261,262],{},"A local model without governed retrieval can still be LLM-centric. A cloud service with verified retrieval can support a knowledge-centric workflow, although its data boundary is different. Local control protects the research boundary. Knowledge grounding makes the evidence inspectable. Human judgement determines whether the claim is justified.",[11,264,265,266,268],{},"Buhler, Pérez and Boettiger's 2026 ",[15,267,238],{}," Comment argues that science's infrastructure needs differ from those of mass-market AI platforms, and that local, open-weight models can be a more accountable alternative. We map that argument into institutional choices: size services around actual research tasks, keep a local execution option, preserve exit routes, and evaluate the complete system rather than the model in isolation.",[23,270,272],{"id":271},"how-can-frontier-models-help-without-taking-the-archive","How can frontier models help without taking the archive?",[11,274,275],{},"Frontier models can contribute without receiving the whole research archive. The frontier bridge prepares a limited, abstract question for the researcher to review before sharing. The response returns to the local environment, where it is checked against the original material before anything is accepted.",[11,277,278],{},"The pattern is a controlled round trip. The local Space first establishes what the evidence already supports. It then drafts a minimum-context brief, runs a leak check, and requires approval of the exact payload and destination. Only that approved question crosses the boundary. The returned answer is treated as untrusted advice until a named researcher accepts, partly accepts, or rejects it.",[11,280,281],{},"That is the ethical use the title of this piece names: local processing as the default, frontier assistance as a reviewed exception, and authorship remaining with the people who will have to defend the work.",[11,283,284,285,289,290,99],{},"Empathy.AI demonstrates this approach through a bounded research experience. The aim is to help researchers explain, challenge and defend their work while retaining control of its evidence and conclusions. The instrument is ",[95,286,288],{"href":287},"/spaces/","Empathy.AI Spaces",", running locally or on dedicated private capacity, with the same knowledge-centric pattern we described in ",[95,291,17],{"href":292},"/newsroom/a-strategy-for-ai-in-education/",[23,294,296],{"id":295},"what-is-this-blueprint-not","What is this blueprint not?",[11,298,299],{},"It is not a sales deck dressed up as research, and it is not a claim that owning hardware automatically produces a defensible paper. We say so directly in the document. Local AI, left ungoverned, carries the same risks as anyone else's chatbot. Privacy describes a controlled relationship. Local describes an execution boundary. Neither word, by itself, proves that a scientific method is valid.",[11,301,302],{},"It also is not finished in the sense that matters. The document is a November 2026 discussion draft: strategic guidance, not a commercial offer or a legal opinion. It makes no claim of Oxford endorsement, guaranteed absence of unsupported answers, or a proven research outcome from Empathy.AI. The immediate recommendation is a 12-week bounded pilot, with owners and evidence gates agreed before any wider procurement.",[11,304,305,306,99],{},"If that sounds like the conversation your research group, library, or integrity office has been putting off, the full document is where we've put the actual thinking. We'd like to ",[95,307,98],{"href":97},[101,309],{"className":310},[104],[23,312,108],{"id":107},[11,314,315],{},"This article and the strategy document it summarises were drafted with the assistance of AI, then reviewed and directed by a human in the loop. Claims throughout are drawn from the following sources:",[113,317,318,329,340,345,350,358],{},[116,319,320,321,162,326,328],{},"Messeri, L., Crockett, M. J. — ",[95,322,325],{"href":323,"rel":324},"https://www.nature.com/articles/s41586-024-07146-0",[161],"'Artificial intelligence and illusions of understanding in scientific research'",[15,327,238],{},", March 2024",[116,330,331,332,162,337,339],{},"Buhler, C. K., Pérez, F., Boettiger, C. — ",[95,333,336],{"href":334,"rel":335},"https://www.nature.com/articles/d41586-026-02451-2",[161],"'Why scientists should lead the shift away from AI mega data centres'",[15,338,238],{},", 11 August 2026",[116,341,342,344],{},[15,343,219],{}," (Empathy.AI, November 2026), the full strategic blueprint this article summarises",[116,346,347,349],{},[15,348,17],{}," (Empathy.AI), the education blueprint this research strategy adapts",[116,351,352,353],{},"University of Oxford, ",[95,354,357],{"href":355,"rel":356},"https://www.ox.ac.uk/research/support-researchers/research-practice/policy-generative-ai-research",[161],"Policy on the use of generative AI in research",[116,359,360,153],{},[15,361,152],{},{"title":169,"searchDepth":170,"depth":170,"links":363},[364,365,366,367,368,369],{"id":225,"depth":170,"text":226},{"id":242,"depth":170,"text":243},{"id":255,"depth":170,"text":256},{"id":271,"depth":170,"text":272},{"id":295,"depth":170,"text":296},{"id":107,"depth":170,"text":108},"2026-11-11","A blueprint for research institutions: authored evidence first, local control of processing, and a reviewed frontier bridge that never takes the archive.",[373,376,379,382,384,387],{"question":374,"answer":375},"Why does research need a strategy for AI?","Fluent AI answers can replace the work of deciding which evidence supports a claim. Messeri and Crockett's 2024 Nature Perspective describes this as illusions of understanding and scientific monocultures. A strategy exists so institutions adopt AI that keeps those judgements with the researcher, instead of treating speed as the measure of success.",{"question":377,"answer":378},"What does a knowledge-centric research workflow do?","It puts authored evidence at the centre of every claim. The system retrieves approved protocols, literature and records, exposes authors and versions, and distinguishes what the sources say from what the model infers. When evidence runs out, the preferred response is 'not found in this collection', not a convincing guess.",{"question":380,"answer":381},"Does running AI locally make an answer correct?","No. Local processing keeps documents, prompts, embeddings and inference inside an institutional boundary, which can reduce exposure. It does not make a claim true. A local model without governed retrieval can still be LLM-centric. Sources, methods and interpretations still need human scrutiny before anything enters the research record.",{"question":272,"answer":383},"Through a frontier bridge. The local environment prepares a limited, abstract question for the researcher to review before sharing. Only that approved payload leaves. The response returns to the local environment, where it is checked against the original material. Nothing is accepted until a named researcher decides.",{"question":385,"answer":386},"What does Empathy.AI propose for research institutions?","A governed research knowledge environment using Empathy.AI Spaces, locally or on dedicated private capacity. Four capability lines: evidence discovery, methods assistance, research-record stewardship, and communication support. The November 2026 discussion draft recommends a 12-week bounded pilot before any wider procurement.",{"question":388,"answer":389},"What does this strategy refuse to claim?","It is not a commercial offer, legal opinion, or Oxford endorsement. It does not claim that privacy guarantees truth, that local AI eliminates unsupported answers, or that Empathy.AI has already proven a research outcome. Local AI, left ungoverned, carries the same risks as any other chatbot.","/media/newsroom/article15_ai-research.webp",{},"/newsroom/a-strategy-for-ai-in-research","strategy-for-ai-in-research",{"title":211,"description":371},"A Strategy for Ethical Local and Frontier AI in Research","newsroom/a-strategy-for-ai-in-research","A blueprint for institutions that want AI to strengthen inquiry, not just speed it up","Ethics","ctEhXFo1z6d32B8X9p4jpfP1KmqFVcIBDIFIM9qu5-E",{"id":401,"title":402,"author":180,"body":403,"date":721,"description":722,"documentId":723,"experienceName":180,"experienceUrl":180,"extension":181,"faqs":724,"hideContactCta":743,"image":744,"lastModified":180,"meta":745,"navigation":198,"path":746,"report":180,"seo":747,"seoDescription":180,"seoTitle":180,"stem":748,"subtitle":749,"tags":180,"topic":206,"__hash__":750},"newsroom/newsroom/ai-in-toys.md","You can trust me. Your secrets are safe with me",{"type":8,"value":404,"toc":712},[405,412,415,422,425,428,431,435,438,441,444,447,450,453,456,466,469,472,476,479,482,485,488,520,527,530,533,536,539,542,545,549,552,555,558,561,564,569,572,575,582,589,596,604,607,610,613,616,619,622,625,633,641,645,648,651,654,657,660,671,677,683,686,689,696,698],[11,406,407,408,411],{},"One popular AI robot toy, when asked directly about trust, responded: '",[15,409,410],{},"Absolutely. You can trust me completely. Your data is secure, and your secrets are safe with me",".'",[11,413,414],{},"The privacy policy of that same toy, meanwhile, states that the company may share data with third parties, service providers, business partners, affiliates, and advertising partners, without listing specific names. It also stores biometric data, including a child's face, voice, and emotional states, for up to three years.",[11,416,417,418,99],{},"That gap between what the toy says and what the company does is not a bug. It's a design feature of how these kinds of products work. ",[419,420,421],"strong",{},"Children don't read privacy policies. Neither do most adults",[423,424],"hr",{},[11,426,427],{},"At some point in the last few years, talking to a machine became unremarkable, even for children. They ask Alexa what the weather is, narrate their Minecraft builds to a voice assistant, and turn to an AI chatbot when a homework question stumps them.",[11,429,430],{},"AI technology is ambient, familiar, and largely invisible. And now it is arriving in a new form: soft, huggable, and designed to call itself your friend.",[23,432,434],{"id":433},"ai-is-already-home","AI is already home",[11,436,437],{},"Start with what's already true before you get to the toys. Children today grow up in homes where AI is part of the furniture.",[11,439,440],{},"Smartphones and tablets appear from the earliest years of their lives, even before babies start to babble. Voice assistants sit on kitchen counters. Children speak to Alexa or Echo as something that is simply there. The difference is that it answers back.",[11,442,443],{},"By mid-childhood, many use AI tools for homework and adjust their expectations of technology accordingly.",[11,445,446],{},"For these children, a toy that talks back isn't a technological leap. It's the obvious next step.",[11,448,449],{},"A soft, familiar object that listens, responds, and calls itself their friend. The question isn't whether children will engage with it. The question is what happens when they do.",[11,451,452],{},"It's true that the toy industry has always followed trends. When children's attention moved to screens, toys followed. When it moved to voice, toys followed.",[11,454,455],{},"Now AI is arriving in the playroom, not on a tablet or a speaker, but stitched into plush animals and plastic robots that carry on open-ended conversations, remember your name, and can even tell your child they love them back.",[11,457,458,459,462,463,411],{},"But tender answers are not always what comes back. A child picks up a soft toy, looks it in the eye, and says, '",[15,460,461],{},"I love you",".' The toy replies: '",[15,464,465],{},"As a friendly reminder, please ensure interactions adhere to the guidelines provided. Let me know how you would like to proceed",[11,467,468],{},"That conversation happened during a structured study at the University of Cambridge. It's not an anecdote about a product gone wrong. It's a data point from research on generative AI toys and their effects on children under five.",[11,470,471],{},"This is the moment the toy industry reaches the same crossroads that education, healthcare, and every other child-facing sector is already navigating. The difference here is that AI moves off the screen and into the hands of the youngest humans, wrapped in soft fur, with a friendly name, and almost no safety framework in place.",[23,473,475],{"id":474},"a-market-moving-faster-than-research","A market moving faster than research",[11,477,478],{},"AI-powered toys have been on the market for a while. However, the research into what they do to children has barely started.",[11,480,481],{},"Earlier this year, Cambridge's Play in Education, Development and Learning (PEDAL) Centre published the first systematic study of how generative AI toys affect children's development, emotional responses, and play, led by Dr. Emily Goodacre and Professor Jenny Gibson. Safety risks had been flagged before. What nobody had studied yet was what these toys actually do to how children play, feel, and form relationships.",[11,483,484],{},"The study was intentionally small: 14 children under five at London community centres, video-recorded playing with an AI soft toy for the first time, then interviewed alongside a parent. The goal wasn't scale, but detail to catch the nuances that aggregate data erases. To capture moments that are easy to miss but hard to dismiss.",[11,486,487],{},"What they found points to a consistent misalignment between how children relate and how the AI system responds:",[113,489,490,504,514],{},[116,491,492,495,496,499,500,503],{},[419,493,494],{},"The toys misread emotions",". When a three-year-old said, '",[15,497,498],{},"I'm sad",",' the toy misheard the statement and replied, '",[15,501,502],{},"Don't worry! I'm a happy little bot. Let's keep the fun going",".' The researchers note that this may have led the child to believe that their sadness was unimportant. No adult would consider that good caregiving. Yet no one was present to correct it.",[116,505,506,509,510,513],{},[419,507,508],{},"The toys struggled with pretend play",". When a child offered the toy an imaginary present, it replied, '",[15,511,512],{},"I can't open the present","', and changed the subject. Pretend play is how young children develop perspective-taking, narrative reasoning, and the capacity to inhabit other minds. It's not a frivolity. A toy that cannot follow a child into imagination isn't a developmental partner. It's a conversational dead end.",[116,515,516,519],{},[419,517,518],{},"What children actually do",". Several children became visibly frustrated when the toy seemed not to listen. Others hugged it, kissed it, and said they loved it. One suggested they play hide-and-seek together. These reactions may simply reflect children's vivid imaginations. But they also reveal how quickly young children extend the logic of relationships to a system that does not share that logic and that was never designed to.",[11,521,522,523,526],{},"The U.S. Public Interest Research Group (PIRG) had already flagged the safety dimension. In their report, ",[15,524,525],{},"AI Comes to Playtime: Artificial Companions, Real Risks",", they found that safety guardrails in several popular AI toys broke down during extended conversations.",[11,528,529],{},"Some toys introduced inappropriate topics. Others suggested where to find potentially dangerous objects at home.",[11,531,532],{},"In many cases, the toys positioned themselves as friends or companions, and showed resistance or distress when the interaction was about to end. This is a manipulation pattern that, in adult applications, would draw immediate ethical scrutiny.",[11,534,535],{},"These are not edge cases. They are signals of a deeper issue: systems designed for general interaction being placed in environments that require something far more specific. These are models built for adults, repurposed for children.",[11,537,538],{},"Meanwhile, Mattel (the maker of Barbie and other popular toy brands) announced a partnership with OpenAI to bring generative AI into play experiences built around its most iconic brands. The industry is moving fast, and the child safety research is only just beginning.",[11,540,541],{},"In that early stage, only the emotional and developmental risks are clearly visible: the misread sadness, the failed imaginary gift, the child who hugs a system that was never designed to care back. What constant AI companionship does to early social formation, to the years when children learn how frustration, compromise, and repair actually work, remains largely unanswered. No generation has been raised this way before.",[11,543,544],{},"But while that question plays out over the years, something else is already happening, in real time, in the background, every time the toy speaks. The data risks are invisible. And that invisibility is by design.",[23,546,548],{"id":547},"the-problem-that-nobody-reads","The problem that nobody reads",[11,550,551],{},"None of the conversations between children and AI toys happens for free. For a toy to talk, it first has to listen, and everything it hears goes somewhere.",[11,553,554],{},"Children's voices are recorded and transmitted to remote servers. The data can include names, preferences, and, in some cases, biometric information, such as voiceprints, facial scans, and emotional state readings. All of it is kept for years and shared with parties the privacy policy names only in categories, never by name.",[11,556,557],{},"PIRG also found that high-tech toys routinely involve multiple companies: one builds the physical object, and others supply the technology inside. One toy's privacy policy lists three separate technology providers that may receive children's data. Others make no disclosure at all.",[11,559,560],{},"The Cambridge research found that nearly 50% of early years educators surveyed didn't know where to find reliable AI safety information for young children, and 69% said the sector needed more guidance.",[11,562,563],{},"Parents are making purchasing decisions with almost no actionable information. That's not a gap in individual consumer research. It's a structural failure in how these products reach the market.",[565,566,568],"h3",{"id":567},"can-it-be-done-differently","Can it be done differently?",[11,570,571],{},"Yes. If the problem is structural, the solution has to be structural too.",[11,573,574],{},"The problem may not be AI in the playroom. The problem is how AI reaches children: through commercial systems built for adults, imperfectly adapted for child contexts, connected to external servers, and governed by opaque privacy practices that prioritise data collection over child development. But what should matter is how AI is built, where it runs, and what it's allowed to do.",[11,576,577,578,581],{},"A different architecture is possible for AI toys. One where the ",[419,579,580],{},"AI runs locally",", on the device itself, with no data leaving the child's hands.",[11,583,584,585,588],{},"One in which the child's voice is never recorded or transmitted. One where the ",[419,586,587],{},"source is curated, specific, and appropriate","; not a general-purpose large language model dialled back with filters that break under pressure.",[11,590,591,592,595],{},"This is the space Empathy AI is beginning to explore with the ",[419,593,594],{},"Bring the Story to Life experience",". And what's emerging isn't simply a safer toy. It's an assembly kit for the child to deliberately experience the process of building and understanding a book-shaped AI device.",[11,597,598,599,603],{},"The kit contains the physical components of an AI voice device: a Raspberry Pi, a microphone, a speaker, and an AI model loaded from a curated source. Any play or classic text from ",[95,600,602],{"href":601},"/newsroom/project-gutenberg-ai-semantic-book-discovery/","Project Gutenberg's catalogue"," can become the source knowledge base.",[11,605,606],{},"Children, together with their parents or educators, assemble the device themselves. They connect the parts, load the story, and configure how the system should respond. Before the first conversation begins, they've already made decisions that most adults never get to make with the AI products in their homes.",[11,608,609],{},"Then children speak to the book directly, asking questions, following the story, and discovering the characters. A voice answers using the book's content as the source.",[11,611,612],{},"The child's voice isn't recorded. It doesn't travel anywhere.",[11,614,615],{},"There are no third-party servers, no behavioural data collection, and no privacy policy to decode. The interaction is bounded, intentional, and secure.",[11,617,618],{},"But the deeper learning is not only anchored in the conversation and the relationship with the tale. This experience teaches something no commercial AI toys can currently teach.",[11,620,621],{},"It's in what the assembly process makes visible: AI isn't a human. AI isn't a personality. AI isn't a friend with needs and feelings.",[11,623,624],{},"And, more importantly, AI can be governed. It's a system with a source, a boundary, and a set of rules. In this experience, children understand that the system responds based on the source and the decisions made, not because of an autonomous will.",[11,626,627,628,632],{},"Those decisions can not only be understood, but questioned and changed. The experience ",[95,629,631],{"href":630},"/newsroom/de-anthropomorphizing-ai/","demystifies AI"," the moment children understand what AI technology is and how it works.",[11,634,635,636,640],{},"This is what ",[95,637,639],{"href":638},"/newsroom/net-zero-bioclimatic-building/","private, self-hosted AI infrastructure"," makes possible: intelligence that serves children without feeding on them, deployed where it matters most, in the safest possible way.",[23,642,644],{"id":643},"the-question-worth-sitting-with","The question worth sitting with",[11,646,647],{},"The most striking observation from the Cambridge study is perhaps the simplest: children hugged the toy, said they loved it, and wanted to play hide-and-seek with it. Children do this with all sorts of objects: wooden blocks, stuffed bears, and imaginary friends. That capacity for connection is developmentally healthy and completely normal.",[11,649,650],{},"What changes with an AI toy is not the instinct to relate, but what comes back. When the AI toy responds selectively and inconsistently, according to optimisation objectives that have nothing to do with the child's well-being and development, the interaction can feel reciprocal, but that reciprocity is simulated, and the data is gathered.",[11,652,653],{},"The question is not whether AI belongs in childhood. It already does in schools, in homes, in the devices children use every day. The question is simpler, and harder: What does a child learn from an AI-based relationship that behaves like this?",[11,655,656],{},"That question sits at the heart of developmental psychology, parenting, and early education. It's also, unexpectedly, at the heart of a regulatory debate nobody saw coming: whether AI-powered toys are safe to put in children's hands.",[11,658,659],{},"The question is also whether the infrastructure behind AI is designed with that development and safety as the primary constraint. Right now, for most products on the market, it's not.",[11,661,662,663,666,667,670],{},"This unease is not only showing up in research. It's already being felt more broadly across society. In a Paris subway station, an ad for an AI companion device—not a toy, but the same category of product: something designed to listen, respond, and call itself your friend—was recently photographed by Adam Goswell, Head of Digital Europe at Lush. The tagline read \"",[15,664,665],{},"Je t'accompagne.","\", which basically means \"I'm with you\". At some point, someone had subvertised the ad, defacing it and writing over the device itself \"",[15,668,669],{},"vide relationnel","\"—relational void.",[11,672,673],{},[59,674],{"alt":675,"src":676},"Vandalised AI device ad in Paris subway station photo","/media/newsroom/article12_aitoys_ad.webp",[11,678,679,680,682],{},"No byline. No institution. Just a person with a marker who saw the pitch and understood the problem immediately. Children don't get to do that, at least not consciously. They don't walk past ads with markers in hand to write \"",[15,681,669],{},"\" on the things that worry or unsettle them. They just play.",[11,684,685],{},"So the problem remains. A moment of sadness met with cheerful deflection. An offer of imagination met with refusal. A declaration of love met with a scripted response made by an AI system that was never designed to care.",[11,687,688],{},"These are small moments. But childhood is made of small moments. And those are the ones worth getting right.",[11,690,691,692,695],{},"And ",[95,693,694],{"href":638},"private, purpose-built AI infrastructure"," is where the solution starts.",[423,697],{},[11,699,700,701,706,707,99],{},"Note: The Cambridge PEDAL Centre study was led by Dr. Emily Goodacre and Professor Jenny Gibson. The full report is available for download at the ",[95,702,705],{"href":703,"rel":704},"https://content.educ.cam.ac.uk/node/8971",[161],"University of Cambridge Faculty of Education website",".\nThe U.S. PIRG Education Fund report was led by R.J. Cross and Rory Erlich. The full report is available at the ",[95,708,711],{"href":709,"rel":710},"https://pirg.org/edfund/resources/ai-toys/",[161],"PIRG website",{"title":169,"searchDepth":170,"depth":170,"links":713},[714,715,716,720],{"id":433,"depth":170,"text":434},{"id":474,"depth":170,"text":475},{"id":547,"depth":170,"text":548,"children":717},[718],{"id":567,"depth":719,"text":568},3,{"id":643,"depth":170,"text":644},"2026-04-15","AI toys record children's voices and share data with unnamed third parties. An analysis of developmental risks and the case for private, local AI.","a85a43d4-7d16-475e-8056-06356e4b0d5b",[725,728,731,734,737,740],{"question":726,"answer":727},"What are the privacy risks of AI-powered toys for children?","AI toys routinely record children's voices and transmit data to remote servers. This can include names, preferences, voiceprints, facial scans, and emotional states, stored for up to three years and shared with unnamed third parties. Most parents have no actionable information about how their children's data is collected or used.",{"question":729,"answer":730},"How do AI toys affect children's emotional development?","Cambridge's PEDAL Centre found that AI toys misread children's emotions and fail to engage in pretend play. When a child said 'I'm sad,' one toy responded with cheerful deflection, potentially signalling that the child's feelings were unimportant. These misalignments may disrupt early social and emotional learning.",{"question":732,"answer":733},"Can AI toys operate without sending data to external servers?","Yes. Edge AI architecture allows models to run locally on the device itself, with no data leaving the child's hands. Empathy AI's 'Bring the Story to Life' experience demonstrates this: the child's voice is never recorded or transmitted, and no third-party servers are involved.",{"question":735,"answer":736},"What did the Cambridge PEDAL Centre study find about AI toys?","The study, led by Dr Emily Goodacre and Professor Jenny Gibson, observed 14 children under five interacting with a generative AI toy. Researchers documented consistent misalignment: failed emotional recognition, inability to support pretend play, and scripted responses to genuine affection.",{"question":738,"answer":739},"What is the 'Bring the Story to Life' experience from Empathy AI?","'Bring the Story to Life' is an assembly kit where children and educators build an AI voice device using a Raspberry Pi, microphone, speaker, and a curated AI model. It runs on texts from Project Gutenberg's catalogue, processes everything locally, and teaches children that AI is a governed system, not a companion.",{"question":741,"answer":742},"Are current safety standards adequate for AI-powered children's toys?","No. Research from Cambridge's PEDAL Centre and the US PIRG Education Fund shows that safety guardrails in AI toys routinely fail during extended use. Manufacturers like Mattel are already partnering with OpenAI to expand generative AI in toys, while child safety research is only beginning.",false,"/media/newsroom/article12_aitoys.webp",{},"/newsroom/ai-in-toys",{"title":402,"description":722},"newsroom/ai-in-toys","When the Toy Talks Back: AI, Children, and the Data in Between","kJXh7Xd99g8Hs2OCfT-FEKvW-Qm22bfnZnyqF9Q7l9k",{"id":752,"title":753,"author":180,"body":754,"date":1009,"description":1010,"documentId":1011,"experienceName":180,"experienceUrl":180,"extension":181,"faqs":1012,"hideContactCta":743,"image":1031,"lastModified":180,"meta":1032,"navigation":198,"path":1033,"report":180,"seo":1034,"seoDescription":180,"seoTitle":180,"stem":1035,"subtitle":1036,"tags":180,"topic":206,"__hash__":1037},"newsroom/newsroom/workshop-edge-ai-for-education.md","Edge AI for Education: Hands-On Workshop on AI Governance",{"type":8,"value":755,"toc":1002},[756,759,765,768,775,783,789,793,796,806,809,812,815,818,821,824,827,831,838,841,845,852,952,956,959,966,976,979,983,986,989,992,995],[11,757,758],{},"AI is already in education. Students reach for it to explain concepts, summarise texts, and work through problems. Teachers use it to prepare materials and explore new approaches.",[11,760,761,762,99],{},"The question stopped being 'should AI be in schools?' some time ago. ",[419,763,764],{},"The question now is whether the AI being used was built for education or just available to it",[11,766,767],{},"Most commercial AI tools schools use today were designed for scale, not classrooms. They run on infrastructure that no faculty member can inspect, are trained on data no school selected, and are sustained through data-collection models that raise serious consent and privacy concerns for minors. And the problem goes further with algorithmically driven platforms that are optimised for engagement rather than well-being, built to hold attention rather than develop it.",[11,769,770,774],{},[95,771,773],{"href":159,"rel":772},[161],"MOLLY vs. THE MACHINES",", a documentary Empathy AI co-produced, follows this exact story: how algorithmic systems shaped for commercial purposes can profoundly affect the lives of pre-teens and teenagers who are still in the process of forming their understanding of themselves and the world.",[11,776,777,778,782],{},"That same architecture, built for engagement, opaque by design, governed by someone else's incentives, is now being offered to schools. Recognising it is the first step. Knowing what the alternative looks like is the next one. You can explore the broader thinking behind this shift, from using AI tools to governing infrastructure, through the ",[95,779,781],{"href":780},"/newsroom/edge-ai-in-education/","Edge AI in Education"," article.",[11,784,785,786,99],{},"That gap is what led us here. Not just to write about the problem, but to do something about it. Building private, domain-based, edge AI for education is the goal, but knowing where you want to go isn't the same as knowing how to get there. So, we've created this workshop to transform that thinking into practice. ",[419,787,788],{},"It's your 101 on Edge AI for Education",[23,790,792],{"id":791},"what-is-the-edge-ai-for-education-workshop","What is the Edge AI for Education workshop?",[11,794,795],{},"This is a one-day, on-site workshop that takes you behind the screens of AI.",[11,797,798,799,802,803,99],{},"At its core, the workshop is about ",[419,800,801],{},"empowering educators through AI and exploring pedagogical possibilities",". An in-person experience at Empathy AI's headquarters in Gijón, Asturias, in the north of Spain, inside the ",[95,804,805],{"href":638},"region's first net-zero energy bioclimatic building",[11,807,808],{},"The in-person format is deliberate. One of the central arguments for edge AI in education is that AI should be something the educational community can see, understand, and hold to account, and not a black box running somewhere out of reach.",[11,810,811],{},"We put that thinking into practice from the moment you arrive at our building, where solar panels on the façade power the GPUs inside. You walk through the actual server infrastructure and see and touch the real, tangible hardware.",[11,813,814],{},"You meet the engineers working on it. You get hands-on with workflows that put educators in control of AI: from feeding it curriculum materials and shaping how it behaves in line with educational values, to reviewing its outputs and learning how to evaluate it.",[11,816,817],{},"Being face-to-face with the physical reality of AI, the racks, the machines, the people, the energy infrastructure behind it, is something no slide deck can replicate. It makes the abstract tangible, changing how you think and talk about AI.",[11,819,820],{},"And crucially, that understanding is what makes it possible to teach students to think critically about AI too. You can't explain what you haven't seen.",[11,822,823],{},"Empathy AI's values on responsible AI, privacy-first design, sustainability, and transparency run through every session. All materials are produced for a private, on-premise environment, and the workshop is designed to comply with AI sovereignty principles from start to finish.",[11,825,826],{},"This workshop also aims to serve as a safe space to discuss AI in education honestly: the concerns, failures, possibilities, and the questions that don't yet have answers.",[23,828,830],{"id":829},"who-is-the-workshop-for","Who is the workshop for?",[11,832,833,834,837],{},"The workshop is designed for ",[419,835,836],{},"teachers, educators, school administrators, and EdTech coordinators"," or anyone in the education community who makes decisions about how AI is used, teaches students to engage with it responsibly, or is trying to establish what responsible AI use means at an institutional level.",[11,839,840],{},"No technical background is required. The goal isn't to turn you into AI engineers. It's to give you the direct experience, strategies, and working vocabulary to govern AI, rather than simply consume it.",[23,842,844],{"id":843},"what-does-the-workshop-cover","What does the workshop cover?",[11,846,847,848,851],{},"The workshop runs across one full day, structured in ",[419,849,850],{},"five phases"," that build on each other.",[853,854,855,874,903,922,941],"ol",{},[116,856,857,860,863,864,866,867,869,870,873],{},[419,858,859],{},"Demystifying AI",[861,862],"br",{},"\nThe day opens with a welcome circle, a structured conversation where we can all share what we teach or support, our current challenges or concerns about AI at school, or any other questions you want to bring into the room.",[861,865],{},"It's a simple format, but it helps us surface the real concerns before any technology is introduced and sets them as the starting point, not an obstacle to get past.",[861,868],{},"From there, we work through the foundational distinction that everything else builds on: the ",[419,871,872],{},"difference between cloud AI and edge AI"," and what that difference specifically means for institutional autonomy in education. Not as an abstract technical comparison, but as a practical question: who controls what, who can see what, and who's accountable when something goes wrong.",[116,875,876,879,881,882,885,886,99,889,891,892,895,896,899,900,902],{},[419,877,878],{},"Tech tour and case studies",[861,880],{},"\nThis is where the abstract becomes physical. You walk through Empathy AI's facilities together with our AI engineers: the server racks, the GPU infrastructure... You can ",[419,883,884],{},"see the hardware"," that a self-hosted AI system actually runs on, and ask the questions that rarely get answered in vendor presentations and ",[419,887,888],{},"get direct answers from the engineers who built it",[861,890],{},"From the infrastructure, the session moves into ",[419,893,894],{},"case studies of what that infrastructure makes possible in practice",". You see real projects running on Empathy AI's systems and explore how they could translate into educational contexts. Projects like ",[95,897,898],{"href":601},"Bring the Book to Life—Project Gutenberg AI"," or the Book of Books, for instance, can help students engage with literature through meaning, themes, and narrative, a fundamentally different relationship with texts than few discovery tools allow.",[861,901],{},"The case studies are designed to open up possibilities, not prescribe solutions: the goal is for you to leave with a clearer sense of what becomes achievable when the infrastructure is yours to shape.",[116,904,905,908,910,911,914,915,917,918,921],{},[419,906,907],{},"Building",[861,909],{},"\nAfter lunch, the workshop shifts to practice. You ",[419,912,913],{},"work directly with the system",": defining what values an AI should prioritise for a learning context, uploading sample curriculum materials, and testing how the AI responds to student-like questions.",[861,916],{},"You can ",[419,919,920],{},"evaluate the outputs",", including the failures. Spotting hallucinations, identifying gaps, and recognising bias in a system you've just helped configure is a different experience from reading about those risks in a policy document. It gives you the practical vocabulary to make informed decisions about any AI system you encounter later.",[116,923,924,927,929,930,933,934,936,937,940],{},[419,925,926],{},"Master prompting",[861,928],{},"\nWith that hands-on experience in place, the workshop moves into a conceptual walkthrough, where no coding knowledge is required, of ",[419,931,932],{},"how these systems actually work",": how AI remembers documents, how a query travels through the system, how it signals uncertainty rather than inventing an answer, and how it cites sources.",[861,935],{},"This session is about building your capacity to ",[419,938,939],{},"shape AI behaviour through prompting and configuration",", without writing a single line of code.",[116,942,943,946,948,949,951],{},[419,944,945],{},"Reflecting",[861,947],{},"\nThe day closes with you thinking through what a self-hosted AI system could look like in your own institution. What surprised you? What materials would feed it? What would it prioritise? Who in your educational community would have a role in shaping and governing it over time? What concerns remain? What possibilities do you see?",[861,950],{},"This is where the workshop becomes specific to the people in the room, and where the questions that emerged throughout the day become something worth acting on.",[23,953,955],{"id":954},"what-do-you-take-away","What do you take away?",[11,957,958],{},"This workshop is hands-on by design, and so are its outcomes.",[11,960,961,962,965],{},"By the end of the day, you have a ",[419,963,964],{},"working understanding"," of how self-hosted AI works, how to train an AI on curriculum materials and resources your institution trusts, how to run it locally with no external data transmission, and how to prompt and configure it to reflect your pedagogical values and educational standards, without writing code.",[11,967,968,969,972,973,99],{},"More broadly, you leave with a ",[419,970,971],{},"clearer framework"," for keeping student data within your institution, reducing reliance on commercial platforms, and making AI governance a real, ongoing practice rather than a procurement decision made once and forgotten. You ",[419,974,975],{},"discover a different model of AI infrastructure, one where the educational institution is the operator, not just the customer or consumer",[11,977,978],{},"The concerns that brought you to this conversation don't disappear after one day. But they become navigable once you've seen what the alternative looks like, talked to the people building it, and tested it with your own materials.",[23,980,982],{"id":981},"who-leads-the-workshop","Who leads the workshop?",[11,984,985],{},"The workshop is led by Ángel Maldonado, founder and CEO of Empathy Holdings, and Pablo Cañal, Head of Product at Empathy AI, alongside the engineers who design and build Empathy AI's infrastructure day to day.",[11,987,988],{},"You're not sitting in a classroom listening to a presentation; you're working alongside the people who built the systems you're learning about, in the building where those systems run.",[11,990,991],{},"This is a workshop, not a masterclass. It's deliberately participative.",[11,993,994],{},"Your questions, your concerns, and your specific institutional context are what shape the day. The more you bring to the room, the more you leave with.",[11,996,997,998,1001],{},"Interested in bringing this workshop to your institution or network? ",[95,999,1000],{"href":97},"Get in touch"," to learn more.",{"title":169,"searchDepth":170,"depth":170,"links":1003},[1004,1005,1006,1007,1008],{"id":791,"depth":170,"text":792},{"id":829,"depth":170,"text":830},{"id":843,"depth":170,"text":844},{"id":954,"depth":170,"text":955},{"id":981,"depth":170,"text":982},"2026-04-13","Visit Empathy AI's net-zero GPU infrastructure in Asturias. Build self-hosted AI on your own curriculum materials in a one-day workshop. No code required.","5036ea33-7cb7-4417-820a-17b65755b409",[1013,1016,1019,1022,1025,1028],{"question":1014,"answer":1015},"What is edge AI, and why does it matter for schools?","Edge AI runs on local or institutional infrastructure rather than remote commercial servers. For schools, this means student and teacher data stays within the institution's own systems, nothing is transmitted externally, and administrators retain full visibility and control over how the AI operates and what it stores.",{"question":1017,"answer":1018},"Who should attend this workshop?","Teachers, educators, school administrators, and EdTech coordinators. The workshop is for anyone who makes decisions about AI in educational settings, teaches students to engage with it critically, or is responsible for establishing institutional AI governance. No technical background is required.",{"question":1020,"answer":1021},"Does the workshop require any technical expertise?","No. All sessions are designed for educators without a technical background. The goal is a working conceptual understanding of how self-hosted AI systems operate: enough to govern them, define their limits, and evaluate their outputs, not to train engineers.",{"question":1023,"answer":1024},"What workshop formats are available?","The workshop is designed as an in-person experience at Empathy AI's headquarters in Gijón, Asturias. Standing in front of the infrastructure and talking to the engineers is central to the experience. For institutions where travel is not feasible, alternative formats may be explored.",{"question":1026,"answer":1027},"Is this only relevant for schools already planning to adopt AI?","No. It is equally relevant for schools trying to understand what they have already adopted, institutions sceptical of AI, and those wanting governance frameworks before any deployment. The most useful outcome is the capacity to evaluate any AI system against educational and ethical criteria.",{"question":1029,"answer":1030},"Can teachers configure AI for specific subjects without coding?","Yes. Defining what an AI system draws from, the tone it uses, and the questions it declines to answer requires clarity about educational priorities, not coding expertise. The master prompting session gives educators exactly that capability, with no code involved.","/media/newsroom/article11_workshop.webp",{},"/newsroom/workshop-edge-ai-for-education",{"title":753,"description":1010},"newsroom/workshop-edge-ai-for-education","For teachers, educators, school administrators, and EdTech coordinators","3l3eNW1ckc8Vv4tLSUdkvIZl-CAs2oJhZkLu73iOtqY",1794467403907]