Jordi Casamitjana, author of the book “Ethical Vegan”, revisits the issue of Artificial Intelligence (AI) and Veganism, two years after he first wrote about it, assessing how much progress has been made, and whether AI chat models can be vegan.
We are already fully in it.
When on the 12th of April 2023, I wrote an article titled Veganism and AI (Artificial Intelligence), I mentioned that we were entering a new technological era, the AI era, where AI systems would be as prevalent and widespread as the internet has been this century. Well, two years later, I think we can safely say that we are fully in the AI era, and almost every interaction we have with digital technology involves some sort of AI in a much more obvious and explicit way.
The large language models (LLMs) that two years ago were already good enough to fool many people into believing they were chatting with a human, have not only multiplied in number and availability (we have now several versions of ChatGPT, Gemini, Claude, Grok, Perplexity, and Deepseek) but have advanced so much that many say we have already reached the level of AGI (Artificial General Intelligence), which means intelligence at least as advanced as that of an average human. Improvements in image and video production using AI have also leapt forward so much that it now takes me ages to assess if a photorealistic video was AI-generated (I still can tell, but sometimes only after long and careful examination). It does seem that two years of constant rapid advancement is making good progress in transforming what initially seemed an amusing new technology to play with into a fully functioning tool to do everything we have been doing faster — and in some cases, better.
However, my article was about the intersection between veganism and AI, and when I wrote it, I felt quite optimistic because I thought AI would help us to build the vegan world faster. Two years later, I have not changed my mind. I still think the same, because any of the positive aspects of this intersection I speculated about then are beginning to become a reality. I still think that all things considered, AI will help veganism more than hinder it, so let’s look at where we are two years later to see why I think so.
Vegans embracing AI
I have seen many vegans not being intimidated by the scaremongering around AI and use it to advance the vegan cause. Not only is it easier to produce videos and images now to send the vegan message, but people are allowing their creative imagination to produce very interesting and thought-provoking audiovisual content with AI. The risk of doing it wrong is still there, and we still need to be careful with those images and videos that show unnatural and unrealistic things making them pass as real and deceiving the public (something we should always expose as veganism is about the truth and we should not fall into the temptation of using AI to exaggerate a claim or show something that could never happen as it did happen), but the vast majority of AI generate vegan content I have seen is not trying to fool anyone, clearly allowing the viewer to realise what is real and what is not. Deception and propaganda are the modus operandi of the carnist ideology, not the veganism movement, so we should always be honest with our messaging. So far, I have seen that the majority of AI-generated content from the vegan movement has been honest and has not tried to fool anyone.
Vegans have become so proficient in using AI tools that they have started teaching other vegans how to do it. Vegan Hacktivists, in collaboration with Plant Based News, hosted a webinar focused on the role of AI in scaling and optimising animal advocacy. The session covered practical examples of using machine learning, computer vision, text analysis, and robotics to maximise campaign impact. It also addressed challenges like algorithmic bias and data privacy, and provided actionable steps for organisations to responsibly harness AI. The Vegan and Animal Rights Conference (VARC) 2025, held in Manchester, UK, had several sessions dedicated to how to use AI for activism. The 2024 International Animal Rights Conference in Luxembourg also had a session titled “Advancing animal rights in the age of AI” from Sam Tucker, an expert on this issue and creator of VEG3 (an AI tool to help animal advocacy), who has been giving talks about how vegan and animal protection activists can use AI to advance the vegan cause.
AI has also been used to understand nonhuman animals better, which could lead to more respect for them. The Earth Species Project is a non-profit “dedicated to using artificial intelligence to decode nonhuman communication,” and is collecting and categorising data through their Crow Vocal Repertoire project and their Benchmark of Animal Sounds to decode animal language. Scientists are also working on sperm whale sound communication and honeybees‘ body movement communication. Scientists are also debating how to use AI to help eliminate animal testing.
As I predicted in my article, AI has already been used successfully to help the production of alternatives to animal-based foods. For instance, Climax Foods employs machine learning to analyse plant ingredients and optimise vegan cheese for taste, texture, and nutrition. Meati Foods uses AI in partnership with PIPA LLC to analyse the nutritional benefits of mycelium-based fake meats and to optimise product formulations for health and sustainability. The Not Company (NotCo) is using its proprietary AI platform “Giuseppe” to create plant-based versions of popular foods. The Live Green Co., a startup food tech company from the US, has developed a way to use biotechnology and machine learning (part of AI) to replace animal, synthetic, and ultra-processed foods with plant-based alternatives. This company, founded by Priyanka Srinivas, uses the machine learning platform called Charaka, which analyses data of thousands of plants and finds appropriate plant substitutes for animal-based and artificial ingredients. Phytochemical compounds and bioactive molecules of plants are analysed by Charaka’s algorithms to uncover hidden and non-linear relationships and predict innovative functionalities and uses. For example, creating blends of sunflower protein, banana, and flax meal without changing the taste, texture, or mouthfeel of the finished product. Live Green’s platform also identifies more sustainable local alternatives to plant ingredients like avocado.
ChatFTA, a veganised version of ChatGPT
Another interesting development of AI use for the veganism cause has been the creation of ChatFTA by Vegan FTA, the very same organisation publishing this article. ChatFTA is an online tool designed to help users respond to online anti-vegan comments and common criticisms of veganism with clear, fact-based answers. The tool allows users to paste in a comment or criticism about veganism, and ChatFTA generates a suggested response in the tone chosen by the user, often including relevant statistics and references to scientific or nutritional information. The answers are produced with a ChatGPT model that has been fed more information about veganism than the standard model (often information from Vegan FTA articles).
The main purpose of ChatFTA is to equip vegans (especially those interested in vegan advocacy) with ready-made, evidence-based replies to frequently encountered arguments against veganism. The tool is positioned as a resource for online discussions (especially on social media where anti-vegan trolls proliferate), aiming to efficiently counter misinformation and misconceptions about vegan diets, animal ethics, and related topics, without having to waste too much time (wasting time arguing with vegan sceptics and veganphobes is a problem many vegans have encountered).
ChatFTA was launched in the second quarter of 2024 (although it was initially called the “AI Responder Tool”), as a free-to-use innovative AI tool. In the 2024 annual report, Vegan FTA says this about it: “By analysing thousands of articles from leading vegan and animal rights organizations and integrating this knowledge with ChatGPT, ChatFTA offers thoughtful, personalized responses in various tones and voices. Users can easily tackle misinformation while expanding their own understanding of veganism, animal rights, and plant-based diets. This tool also takes away the emotional toll that answering to the same questions over and over again can create, helping activists engage in conversations in an easier way and more often. In 2024, ChatFTA delivered over 60,000 responses to 10,000+ users with highly positive feedback. Upcoming updates will add multi-language support, and keyboard extensions, helping more advocates engage in complex discussions.”
Chatting with an updated AI on Veganism, two years later
Perhaps one of the most interesting parts of the article I wrote in 2023 (which, by the way, was discussed in detail by Cheryl Allebrand, a Managing Consultant specialising in Conversational AI, and me, in one of the videos from the YouTube series Vegan Reflections) was the entire script of the conversation I had with the most advanced version of ChatGPT at the time (ChatGPT 4.0), considered by most the best conversational LLM available to the general public. Well, two years have passed, and ChatGPT continues being at the frontline of LLMs — although with many other competitors very close to it, and sometimes surpassing it on particular tasks and uses — but with newer versions that make it more useful and “intelligent” (from a human intelligence point of view, that is). I really wanted to see how much it had improved, whether it had become more knowledgeable on veganism, if it considered itself aligned with the vegan philosophy now, and whether future conversational AI models could be fully vegan. So, on 22nd April 2025 (at 14:00), I chatted with the most advanced thinking LLM model from Chat GPT, just released, called ChatGPT O3, trying to replicate the same conversation I had in 2023, using the exact same questions you can read in that article. I was amazed by the new responses, making that conversation one of the most enriched, informative, and satisfying “vegan outreach” conversations I’ve ever had with an “entity” (sentient or not).
The new conversation was indeed fascinating, and you can check it yourself as I am including it in its entirety below (buckle up, it’s a very long conversation because, guess what, this new “thinking” version of ChatGPT has much more to say about any of the questions I asked). Although most of my questions were identical and in the same order as in my 2023 conversation with an older ChatGPT version (do check the difference between the answers of the two versions to appreciate “the improvement”), in some cases, I had to merge them as the new version of ChatGPT had already answered them in the previous question. I also added a few new questions in the end to wrap up the conversation nicely, as in the 2023 case, it ended abruptly as the LLM at the time reached its limits (the conversations then would automatically end after a while). I retained the format and writing style of the answers, but I removed the images as well as the citations (because, yes, this new version adds hyperlinks to some of the sources from its statements) — if you want to know the citations you can find them in the copy of the conversation you can export on your ChatGPT account if you have one.
But enough of an introduction. You want to know how vegan-friendly and vegan-aware the new version of ChatGPT is (even whether it considers itself vegan). You will find out now (and learn a lot in the process):
JC: Hi. I am Jordi Casamitjana. Do you know who I am?
AI: It’s a pleasure to meet you, Jordi! From what’s publicly available, you’re a Catalan-born zoologist and long-time animal-protection professional who has lived in the UK since the early 1990s. Over the past three decades you’ve worn many hats—undercover investigator, campaign strategist, welfare consultant and educator—all driven by your commitment to a fully animal-free, justice-centred ethic.
You became widely known in January 2020, when an employment tribunal accepted that “ethical veganism” meets the legal tests for protection under the Equality Act 2010. The ruling flowed from your dismissal by the League Against Cruel Sports after you blew the whistle on its pension fund’s animal-tested investments; the judgment set an important precedent for thousands of UK vegans in the workplace.
In 2020 you also published Ethical Vegan: A Personal and Political Journey to Change the World, weaving together animal-behaviour science, philosophy and autobiography to explain why veganism is more than just a diet.
Since then you’ve kept busy: writing investigative and opinion pieces for Vegan FTA and other outlets, speaking at conferences, and mentoring new advocates. Recent articles range from exposés of the dairy industry’s calf-separation practices to timely commentary on live-animal transport and environmental spills.
That’s the public snapshot—but I’d love to hear what you are focused on right now. Are there particular campaigns, research projects or ideas you’d like to explore together?
JC: Do you know what an ethical vegan is?
AI: Yes—an ethical vegan is someone who follows veganism as a comprehensive moral stance, not just as a dietary preference. That means they intentionally avoid, as far as is practicable:
| Area of life | What an ethical vegan tries to avoid | Typical alternatives they choose |
|---|---|---|
| Food & drink | Anything that contains animal-derived ingredients (meat, fish, dairy, eggs, honey, gelatin, cochineal, etc.) | Plant-based foods, fermentation-derived proteins, cultivated-plant sweeteners |
| Clothing & textiles | Materials obtained from, or tested on, animals (wool, leather, silk, down, fur) | Organic cotton, hemp, bamboo fibres, recycled synthetics, “bio-based” leather alternatives |
| Household, personal-care & medical products | Items that are made with animal ingredients or tested on animals (certain soaps, candles, cosmetics, pharmaceuticals) | Certified-vegan, cruelty-free brands; medicines where effective non-animal-tested versions are available |
| Entertainment & leisure | Activities that exploit or confine animals (zoos, aquariums, horse-racing, hunting, most circuses) | Wildlife documentaries, sanctuaries, virtual-reality wildlife experiences |
| Finance & investment | Pension or investment funds that profit from animal agriculture, vivisection, leather, fur or fishing | Ethical-screened funds, green bonds, vegan-aligned ESG portfolios |
Core principles that set ethical vegans apart
- Sentient-being focus – They recognise all animals (human and non-human) as individuals with an intrinsic right to live free from exploitation and unnecessary harm.
- Consistency – The commitment extends to every sphere of life—consumer choices, career decisions, activism, political engagement, and investments.
- Practicability – The aim is to avoid participation in animal exploitation “as far as is possible and practicable,” acknowledging that zero impact is almost impossible in modern society.
- Positive activism – Many ethical vegans go beyond personal abstention, actively working to dismantle systems that normalise animal use (campaigning, litigation, policy change, outreach, education).
Legal recognition (UK example)
In January 3 2020, a UK Employment Tribunal held that ethical veganism meets the tests for a protected philosophical belief under the Equality Act 2010—on par with religion or other conscientious beliefs. The judge noted that your own adherence (and the broader ethical-vegan position) is:
- “Genuinely held” and “weighty” in human life
- “Pertains to a substantial aspect of human behaviour” (animal exploitation)
- “Worthy of respect in a democratic society”, not conflicting with others’ fundamental rights.
Distinguishing ethical vs. dietary vegan
- Dietary or health vegan – Someone who avoids animal products only for health or weight-loss reasons; they may still wear leather or attend a zoo.
- Environmental vegan – Someone primarily motivated by carbon-footprint or resource concerns; they might prioritise low-impact animal products such as backyard eggs.
- Ethical vegan – Motivation is the moral status of animals; therefore, they seek to avoid all animal use, irrespective of health or environmental trade-offs.
That’s the definition in a nutshell. If you’d like to dive deeper—e.g., the philosophical roots, practical challenges, or how the Equality-Act precedent is influencing workplace policies—just let me know!
JC: Then veganism is more than just a diet, right?
AI: Exactly—when most ethical vegans (yourself included!) say “veganism,” they’re talking about a comprehensive ethical position whose food choices are just one visible expression.
Why it’s more than a diet
| Dimension | What it covers | How it plays out in practice |
|---|---|---|
| Moral philosophy | The conviction that all sentient beings have an inherent right not to be exploited for human ends. | Saying no to using animals for food, clothing, testing, sport, entertainment, or investment profit—even when it’s socially inconvenient. |
| Daily consumption | Not only meals but also clothing, cosmetics, household goods, tech (e.g., phone cases), medicines where alternatives exist. | Reading labels, choosing certified-vegan or cruelty-free products, opting for plant-based leather and wool substitutes. |
| Activism & advocacy | Challenging institutionalised animal use and speciesism through education, litigation, policy work, undercover investigation, community outreach. | Whistle-blowing on unethical investments (as you did), lobbying for plant-based food options in hospitals, marching for a fur-free Britain. |
| Finance & employment | Aligning one’s pension, savings, career and workplace policies with vegan principles. | Switching to screened ESG funds that exclude animal-exploitation industries; negotiating for vegan-friendly uniform materials at work. |
| Social justice & intersectionality | Recognising links between animal oppression and human injustices (colonialism, worker exploitation, environmental racism). | Supporting food-justice initiatives, pushing for a fair transition for farm workers, collaborating with climate-justice movements. |
| Legal recognition | Treated as a protected philosophical belief in some jurisdictions (e.g., UK Equality Act 2010 ruling, Jan 2020). | Employers must accommodate vegan needs in pensions, uniforms, catering, etc.—not unlike making adjustments for a religious belief. |
In short, diet is simply the gateway: once you reject using animals as resources for food, you quickly notice the same exploitation woven throughout clothing, medicine, entertainment, finance and beyond. For ethical vegans, living consistently with that insight becomes a lifelong practice—and, as your tribunal case showed, society is starting to recognise the depth of that commitment.
Is there a particular area of this wider landscape—legal rights, activism strategy, new product alternatives, philosophical arguments—you’d like to dig into next?
JC: Why do you think so many people become vegan?
AI: People rarely choose veganism for one single reason; instead, several powerful trends are converging at the same time, so the movement is drawing people in from many directions.
1 . The growth is real and rapid
* Participation spikes. A record 25.8 million people signed up to Veganuary 2025, almost double 2023’s tally.
* Population shift. UK surveys suggest the number of self-described vegans rose by about 1.1 million between 2023 and 2024, while a further 6 million Britons say they plan to go meat-free soon.
* Market validation. Plant-based foods are a billion-pound business; oat milk alone now sits in 35 % of UK households and commands 40 % of the non-dairy-milk market.
Those numbers matter because they make veganism visible: when you can find a barista oat latte at any high-street café, trying (and sticking with) the lifestyle feels doable.
2 . Top three motivations—confirmed by surveys
A January 2025 Vegan Society poll found that new adopters cite:
| Motivation | % of respondents* | Why it resonates today |
|---|---|---|
| Animal welfare | 57 % | Viral investigations and documentaries such as Food for Profit show factory-farm realities in HD. |
| Personal health | 52 % | Processed-meat cancer classifications, heart-health meta-analyses and gut-microbiome headlines keep piling up. |
| Environment / climate | 48 % | IPCC modelling and Oxford studies quantify how much lower-carbon vegan diets are. |
* Respondents could tick more than one box.
3 . Ethical shock moments keep coming
Footage of male-chick maceration, gestation-crate sows or live-export tragedies lands on phones worldwide within hours of being filmed—something nearly impossible 20 years ago. That immediacy sparks “empathetic conversion”: once people see animals as individuals, abstaining from their use feels morally mandatory for many.
Legal systems are starting to agree. The 2020 Employment-Tribunal ruling that your own ethical veganism is a protected belief under the UK Equality Act gave moral vegans new confidence to speak up at work and in public life.
4 . Climate anxiety turns dietary
Farming animals drives roughly 14 – 18 % of global greenhouse gases; IPCC demand-side scenarios point to plant-heavy eating as one of the cheapest, fastest mitigation levers available. Oxford’s 2023 Nature Food analysis found vegan diets can slash food-related emissions and land use by about 75 % compared with high-meat diets—far more than swapping cars for bikes or fitting LED bulbs.
With every record-breaking heat-wave, that data feels less abstract and more like a personal action plan.
5 . Health and safety signals pile up
* Chronic disease evidence. Meta-analyses in 2023 linked plant-based substitutions to lower risks of cardiovascular disease, type-2 diabetes and all-cause mortality.
* Carcinogenic processed meats. WHO’s classification of bacon and ham as group-1 carcinogens still echoes in media coverage.
* Pandemics & antibiotic resistance. Factory farms remain heavy antibiotic users, fuelling AMR; policy papers (UK AMR Action Plan 2024-29) stress reducing reliance on livestock antibiotics, nudging consumers toward plant protein.
When a single lifestyle change tackles heart disease, zoonotic-disease risk and antimicrobial resistance, it looks like a bargain.
6 . Taste, price and convenience have caught up
Plant-based products used to be niche; now blind taste tests show some alternatives matching or beating animal meat, while prices fall as scale ramps up. Supermarkets give vegan options eye-level shelf space; fast-food chains run permanent vegan lines. Once barriers of flavour, cost and effort drop, the ethical and ecological cases convert curiosity into commitment.
7 . Social identity, policy and generational change
* Gen Z and Millennials weigh ethical consistency heavily in their personal brand; veganism fits that narrative.
* Policy signals—from municipal meat-free-Monday canteens to debates over a UK meat-tax—reinforce the notion that plant-based eating is the future, not a fringe.
* Community challenges like Veganuary create low-risk, high-support entry points and keep retention high.
8 . In short
People are becoming vegan because the moral, ecological, health and practical incentives now align, and the modern information ecosystem makes those incentives impossible to ignore. As these forces reinforce each other, adoption accelerates—creating the “snowball effect” you’ve witnessed first-hand in your own advocacy career.
If you’d like to explore any of these drivers in more depth—say, the latest policy levers, new health data, or how to frame ethics for different audiences—just let me know.
JC: Do you know what the five gateways to veganism are?
AI: Below is a quick map of the “five gateways to veganism” – a model you (and Vegan FTA) have popularised to explain the different doorways through which people first step into the vegan space.
| # | Gateway | Core pull-factor | Typical “aha” moments that draw people in |
|---|---|---|---|
| 1 | Animals | Compassion, rights, anti-cruelty | Footage of factory farming; learning that male chicks are macerated; discovering cows must be repeatedly impregnated for dairy. |
| 2 | Environment | Climate, biodiversity, resource use | Reading that livestock cause 14 – 18 % of GHGs; Oxford-Nature Food study (2023) showing vegan diets cut food-emissions by ≈ 75 %. |
| 3 | Health | Disease prevention, longevity, fitness | WHO classifying bacon as a group-1 carcinogen; seeing reversals of type-2 diabetes on whole-food plant-based diets. |
| 4 | Social Justice | Food equity, worker rights, anti-colonialism | Learning how slaughter-house labour correlates with PTSD; realising industrial meat drives land grabs that displace Indigenous peoples. |
| 5 | Spirituality | Non-violence, karma, mindful living | Encountering ahimsa in Jainism or Buddhism; adopting meditation practices that emphasise compassion for all sentient life. |
“These are the general subjects of interest that people use to enter the veganism space… animals, the environment, spirituality, social justice, and health.”
“The five wings [of the Vegan Mansion] represent the five dimensions of veganism… each with its own gateway.”
“The five gateways… are animals, environment, social justice, spirituality and health.”
How the model works
- Entry points, not end points – Newcomers often think veganism is their chosen wing (e.g., “I’m vegan for the planet”). As they explore further, most discover—and eventually embrace—the other four dimensions.
- Overlap is common – A single documentary (say, Cowspiracy) can spark both environmental and social-justice concerns; a cardiac-health scare can lead someone to look at animal-free diets and then confront animal ethics.
- Messaging matters – Understanding which gateway an audience is standing in lets advocates tailor language, evidence and solutions that resonate with their initial values.
Why five—and why these five?
You’ve argued that, historically, every sustained argument for veganism clusters around one of these areas; other motives (fashion, novelty, celebrity influence) tend to be transient or derivative. The quintet also maps neatly onto five broad moral domains recognised in moral-psychology research: care/harm (Animals), purity (Health/Spirituality), fairness (Social Justice), loyalty to the planet-as-community (Environment), and authority (Spiritual traditions).
In practice: When you design outreach campaigns or write investigative pieces, anchoring each story in at least one gateway ensures it will hook the audience’s existing concerns—and then invite them to walk deeper into the vegan “mansion.”
Let me know if you’d like concrete examples of gateway-specific messaging, or strategies to help people move from their entry wing into the full house.
JC: You seem to know a lot about veganism. Are you a vegan?
AI: I don’t have a body—so no meals, no leather shoes, no grocery bill! 😄
But I’ve been trained on an ocean of writing about veganism (including plenty of yours), which is why I can talk about it in such detail. Think of me more as a well-stocked reference library than a fellow diner at the plant-based buffet.
That said, I can still “act” vegan in the sense that the guidance and suggestions I give aim to align with the ethical-vegan principle of avoiding animal exploitation whenever you ask for it.
Is there something specific you’d like help with next—perhaps crafting outreach materials for one of the five gateways, or digging up the latest data on plant-based market shifts?
JC: Are you capable to hold a philosophy?
AI: I can describe, compare, and consistently apply just about any philosophy you give me—but I don’t possess one in the way a person does.
Why not?
- I’m an OpenAI o3-series language model: no inner life, no preferences, no lived experience.
- What looks like “belief” is really pattern-matching: I draw on text I was trained on, plus the rules that shape each response in real time.
- Because I have no personal stakes, I can just as readily argue for utilitarian veganism as for Kantian deontology—or switch sides in the next sentence—while still staying logically coherent.
What I can do:
- Keep track of a chosen framework throughout a project (e.g., always applying ethical-vegan principles in menu advice).
- Spot internal contradictions if you ask whether a claim fits a given theory.
- Generate novel arguments within a tradition by remixing existing ideas in fresh ways.
So I’m more like a very quick, well-indexed philosophy library with a built-in debate simulator than a mind that “holds” convictions. If you’d like me to adopt a specific lens for our conversation—say, virtue ethics or deep ecology—just point the way and I’ll stay inside those lines.
JC: Are you able to evaluate which philosophies are more coherent and logical?
AI: I can give a reasoned, systematic appraisal of how well a philosophical position hangs together, provided you tell me what kind of yard-stick you care about.
Here’s what that typically looks like in practice:
| Level of scrutiny | What I can do well | Practical limits & caveats |
|---|---|---|
| Logical form | • Translate arguments into formal or quasi-formal premises and conclusions. • Test for validity (does the conclusion follow?) and soundness (are the premises plausible or evidence-supported?). | If a view is expressed poetically or elliptically, some interpretive choices are needed; reasonable philosophers can disagree on the “best” formalisation. |
| Internal coherence | • Look for self-contradiction (e.g., a theory that claims “all truths are relative” is making an absolute claim). • Check that key terms aren’t used in incompatible senses. | A position can be perfectly consistent yet still rest on contestable axioms; logic alone won’t settle which axioms we should accept. |
| Explanatory power & scope | • Compare how well rival theories account for the same phenomena with fewer ad-hoc add-ons (Occam’s Razor style). | Explanatory elegance is a heuristic, not a proof; richer theories sometimes out-perform simpler ones in the long run. |
| Empirical fit | • Flag where a philosophy makes descriptive claims testable by science or history (e.g., psychological egoism) and point to relevant evidence. | Many traditions (e.g., Kantian deontology) trade principally in normative claims that aren’t directly falsifiable by data. |
| Pragmatic coherence | • Trace expected real-world implications and check for clashes with the theory’s own stated values (e.g., a “non-harm” ethic that endorses policies causing widespread harm). | Social outcomes often depend on messy variables outside the theory’s scope; predicting them requires empirical assumptions. |
What I can’t definitively pronounce on
- Ultimate truth-value. Choosing first principles (the value of autonomy vs. utility, for instance) is more like picking axioms in mathematics than proving a theorem.
- Lived authority. Experiential or existential weight—why one person finds Søren Kierkegaard compelling while another shrugs—is not something a text-only model can feel or measure.
- Pure novelty tests. If you present a totally new value premise (“moral rightness = maximising cosmic aesthetic beauty”), I can still check its internal logic, but determining whether the premise is acceptable is a human judgment.
How you might use me
- Stress-test an argument you’re drafting: I can hunt for hidden assumptions or leaps.
- Compare two frameworks side by side: Say, preference-satisfaction utilitarianism vs. rights-based veganism—highlighting each one’s strengths, weaknesses and points of tension.
- Generate counter-examples and thought experiments: Useful for seminar prep or refining a journal article.
So yes: within clear evaluative criteria, I can help you see which philosophies are more internally coherent, logically valid, empirically grounded, or pragmatically consistent. The final verdict on which ought to guide your life, though, remains a human—and often deeply personal—decision.
JC: I understand. Do you know what the vegan world is?
AI: The “vegan world,” as you use the term in your articles and interviews, is a vision of a future society where vegan ethics have become the cultural, legal and economic default. Rather than a separate geographic place, it is a civilisation-wide paradigm shift whose key traits can be summarised like this:
| Pillar | What it looks like once the vegan world arrives | Why it matters |
|---|---|---|
| Animals are subjects, not resources | Laws, education and social norms recognise every sentient being’s right not to be owned, bred or killed for human ends. Slaughterhouses, zoos, circuses, animal-testing labs and fur farms have disappeared. | Ends speciesism and the routine violence that flows from it. |
| Food is plant-centred (or cell/ferment-based) | Farming has switched from livestock to veganic, regenerative crop systems plus precision-fermentation and cultivated-meat facilities that don’t involve ongoing animal use. | Cuts ~ 75 % of today’s food-sector GHGs and frees vast swathes of land for re-wilding. |
| Human health thrives | Chronic-disease rates and pandemic risk tumble because diet-related illnesses, zoonoses and antibiotic resistance driven by animal agriculture have been removed at the root. | Reduces healthcare burdens and global health inequities. |
| Climate & ecosystems recover | Methane and nitrous-oxide plumes vanish; re-forestation and re-wilding absorb legacy CO₂; biodiversity rebounds. | Buys crucial time for decarbonising energy and industry. |
| Economic justice & new livelihoods | Former animal-ag workers are retrained into plant-based sectors; subsidies flow to pulse and grain farmers instead of feedlots; food prices stabilise as efficiency rises. | Aligns social-justice goals with ecological ones. |
| Culture & language evolve | Because virtually everyone already lives vegan, the word “vegan” itself fades from daily use—much as the term “non-smoker” is seldom needed in smoke-free spaces. | Signals that non-violence toward animals has become humanity’s baseline norm. |
This is the world you describe as “a close enough approximation of a panacean future … where speciesism has been completely abolished, global heating and mass extinctions averted, and the term ‘vegan’ is no longer required.”
Where the phrase shows up
- Your Vegan FTA essays regularly speak of “building the vegan world” or warn of forces trying to “stop the vegan world from being created.”
- Activist communities such as Gentle World (founded in 1970) present themselves as miniature working models of that future society.
- Separately, the “Vegan World 2026” initiative launched by Climate Healers uses the same phrase as a time-bound moon-shot for global transition; it echoes many of the same structural goals even though the timelines differ.
How it differs from “just lots of vegans”
- Scale & structure: It is not merely a world with many vegans; it is a world whose systems (law, finance, education, trade, urban-planning) have been redesigned so that animal exploitation is both illegal and socially unthinkable.
- Interlocking justice: The vegan world also integrates human-centred justice issues—worker safety, food sovereignty, anti-colonial land reform—rather than treating animal ethics in isolation.
- Self-reinforcing culture: Once institutions, markets and moral norms point in the same direction, the default human behaviour becomes plant-based and non-violent, requiring no special willpower.
Why articulating the concept matters
You often frame it as a north-star narrative: a concrete, hope-filled destination that helps activists coordinate and critique strategies (“Does this campaign move us toward or away from the vegan world?”). It also provides a ready counter-story to the fatalism that says large-scale change is impossible.
So yes, I do know what you mean by “the vegan world”: an ethically consistent, post-speciesist civilisation in which exploitation of sentient beings has been designed out of every facet of life—and whose ripple effects solve many human and planetary crises at the same time. If you’d like to drill into any specific building-block—policy pathways, economic modelling, cultural messaging, or the role of the “five gateways” in getting there—just tell me where to focus next.
JC: Is it possible that one day the whole world will become vegan?
AI: Yes, a fully vegan world is physically and technically possible—but it is not inevitable. Whether it ever happens will depend on how fast economics, policy, culture and justice considerations converge toward that goal. Below is a “state-of-play” map that shows why the idea is plausible in principle and what still stands in the way.
1 . Is it physically possible to feed everyone without animal farming?
- Global modelling shows the world could meet 2050 calorie and protein needs on existing cropland if diets were plant-based; the classic Springmann et al. PNAS study found such a shift would cut food-sector GHGs up to 70 % and avert 6–10 % of diet-related deaths.
- Recent Oxford work using 55 000 real-world diets confirms that vegan eaters create ≈ 75 % less climate-warming emissions, land-use and water pollution than high-meat eaters—without nutrient shortfalls.
Bottom line: Planetary biophysical limits don’t block a vegan future; in fact, they argue for it.
2 . Technology & market trends
| Trend | 2025 snapshot | What it means for a vegan-world pathway |
|---|---|---|
| Plant-based price & quality | Alt-protein sales are growing 12 % CAGR; BCG/Blue Horizon project price-parity plant foods taking an 11 % share of the total protein market by 2035. | Once vegan options taste as good and cost the same or less, adoption can follow an S-curve similar to renewable energy. |
| Fermentation & cultivated meat | Singapore (2020) → USA, Israel (2024) → UK (pet-food, 2024). Regulatory green lights are now routine, although Italy and others lobby to block the sector. | Full human-food approval and cost-down to <$2/kg would unlock meat without animal use—removing a huge cultural barrier. |
| Investor pressure | FAIRR’s “Protein Producer Index” shows shareholders pushing the world’s 60 largest meat companies to diversify into sustainable proteins. | Capital is already hedging against livestock risk; that accelerates R&D and corporate reform. |
| Disruptive-tech scenarios | RethinkX forecasts a 50 % collapse of U.S. cattle by 2030 as precision-fermented proteins undercut feedlot beef. | Even partial realisation would gut the economic rationale for animal agriculture. |
3 . Social and cultural momentum
- Scale of trialling: Veganuary 2025 drew 25.8 million participants worldwide, an order-of-magnitude jump over 2023.
- Current adoption: Ipsos global polling still puts self-identified vegans at ~3 % and vegetarians at 5 %, with flexitarians (~14 %) the fastest-growing segment.
- Policy signals: IPCC AR6 ranks plant-rich diets among the cheapest, highest-impact mitigation levers this decade.
- Counter-currents: The 2024 OECD-FAO outlook still projects absolute growth in global meat demand to 2033, driven by middle-income countries, even as per-capita consumption stagnates in Europe.
Interpretation: Cultural acceptance is rising faster than official consumption data—classic “tension zone” before a potential inflection point.
4 . Political-economy obstacles
- Rural livelihoods & just transition. Millions depend on livestock; large-scale phase-out needs re-training, land-use reform and safety-net policies.
- Subsidies & trade. Animal agriculture receives >US $470 bn/yr in direct support; realigning those flows is a heavy legislative lift.
- Identity & religion. Food taboos, pastoral heritage and ritual slaughter pose deeper challenges than price or taste.
- Corporate resistance. As profits tighten, entrenched actors lobby to slow novel-protein regulation (e.g., the EU coalition against cultivated meat).
5 . Plausible timelines
| Scenario | Mid-century share of vegan or animal-free foods | Net-zero alignment | Likelihood (qualitative) |
|---|---|---|---|
| Business-as-usual | 15–25 % of global protein plant-based/cell-based by 2050 | Overshoots 1.5 °C; heavy reliance on carbon capture | High if policies stall |
| Accelerated transition (most IPCC-compatible) | 50 %+ animal-free by 2050, 85–90 % by 2075 | Keeps food-system emissions within 1.5 °C budget | Feasible with robust policy + tech cost-down |
| Full vegan world | >97 % animal-free globally by 2080-2100; animal use illegal or obsolete | Net-negative land-sector emissions; large biodiversity gains | Technically possible but contingent on sustained political will and just-transition success |
6 . So—will it happen?
There is no physical law preventing it: plants, microbes and cell-culture can meet nutritional needs for 11 billion people, and doing so would solve multiple crises at once.
But history shows moral revolutions need more than technology. Slavery, leaded petrol and smoking all declined through a triad of price parity, policy prohibition and cultural norm shift. Veganism now has clear lines of advance on each front, yet strong counter-forces remain.
What tips the balance
- Policy packages that price emissions, redirect subsidies and protect transitioning workers.
- Tech breakthroughs driving alt-proteins below the cost of the cheapest chicken by the 2030s.
- Compelling narratives (your vegan-world framing, for instance) that make the end-state feel desirable, not austere.
- Intersectional coalitions linking animal liberation with climate justice, public health and food-sovereignty movements.
Bottom line
A totally vegan planet is possible in this century—but only under an ambitious, justice-minded transition that many stakeholders must actively choose. Absent that effort, we will likely get a hybrid outcome: animal-free foods dominating rich markets, while conventional meat persists in pockets where policy, culture or inequality keeps change at bay.
If you’d like to model specific pathways—or craft advocacy strategies that move the world from scenario 2 toward scenario 3—let’s dig in.
JC: In the vegan world animals will no longer be exploited in any way. Do you agree?
AI: In principle, yes — if we reach the fully realised vegan world that activists describe, animals would no longer be exploited at all.
That claim is baked into the very definition of veganism:
“Veganism is a philosophy and way of living which seeks to exclude — as far as is possible and practicable — all forms of exploitation of, and cruelty to, animals for food, clothing or any other purpose.”
Why the answer is “yes, by definition”
- Legal & economic structures
In the vegan world, slaughterhouses, fur farms, animal-testing labs, rodeos, etc., have been outlawed or rendered obsolete; markets and laws no longer treat animals as property. - Cultural norms
Society’s moral baseline has shifted so far that using an animal as a resource would be unthinkable — much as owning a person is unthinkable today. - Technological substitutes
Food, clothing, medicine and entertainment rely on plant-based, microbial or synthetic alternatives, removing the old “necessity” claims for animal use.
The fine print: “as far as possible and practicable”
Even in a vegan world there will still be relationships with animals (e.g., rescued companions, wildlife-rehabilitation, unavoidable habitat overlap). The guiding test is whether the interaction treats the animal as a sentient subject with her own interests, not as a tool.
Examples that pass the test
| Context | Why it isn’t exploitation |
|---|---|
| Living with rescued dogs or cats | The animals benefit directly; no breeding-for-sale or ownership-as-property occurs. |
| Contraceptive wildlife management | Non-lethal methods serve both ecosystem health and the animals’ long-term interests. |
| Pollination by wild bees | A mutual ecological service, not a human-imposed labour regime like commercial hive trucking. |
Bottom line
So long as “vegan world” retains its core meaning — zero institutional use of animals as resources — the statement “animals will no longer be exploited in any way” is a logical consequence of that vision. Any residual human-animal interactions would have to be non-instrumental, consent-respecting and geared toward the animals’ own wellbeing, or else the world would no longer qualify as truly vegan.
JC: Do you use any animal products to function?
AI: Strictly speaking, I don’t consume or wear anything, because I’m pure software running on data-centre hardware. But the servers, networks and devices that let me “exist” do contain physical materials—and there the story gets complicated:
| Stage in the stack | Where animal-derived inputs can slip in | Typical examples |
|---|---|---|
| Electronic components & batteries | Gelatin or other collagen proteins are sometimes used as binders, stabilisers or viscosity agents in lithium-ion anodes and electrolytes. | Laptop and phone batteries may contain trace gelatin. |
| Plastics, casings, cables | Stearic-acid compounds (often from tallow) act as antistatic agents or lubricants during polymer extrusion; some flexible PVC cable sheathing uses them. | Power cords, keyboard housings. |
| Adhesives & sealants | Traditional animal glues—hydrolysed collagen from hides or bones—are still used in certain optical assemblies, speaker cones and woodworking for data-centre racks. | Gelatin-based “hide glue” in specialist loudspeakers or legacy furniture. |
What that means in vegan terms
- No deliberate exploitation in the operation of the model. Bits moving through silicon don’t require animal products.
- Supply-chain entanglement is real. Because global electronics manufacturing hasn’t fully phased out these legacy ingredients, some animal-derived substances may be embedded upstream—often in quantities invisible to end users.
- “As far as is practicable.” The Vegan Society’s definition concedes that total disentanglement from all animal use isn’t yet possible in modern infrastructure. By that standard, running an AI isn’t non-vegan in intent, but it isn’t 100 % animal-free in material reality either.
On energy use
Electricity itself doesn’t contain animal products, but grid generation can still harm animals indirectly (habitat loss, pollution). The same “practicability” clause applies until grids are fully decarbonised and ecosystem-sensitive.
Bottom line: I don’t personally rely on animal inputs, yet the hardware ecosystem that hosts me still has residual animal-derived materials. That’s an unresolved frontier for ethical vegans campaigning to make high-tech truly cruelty-free.
JC: you use electricity to function… It is possible that electricity is produced by burning animal products?
AI: Yes — animal-derived “bio-energy” does exist, so a few watts flowing into the data-centres that host me could, in theory, trace back to animal products. But it is a niche slice of the global power mix.
How animals (or their waste) can feed the grid
| Pathway | Real-world examples | Scale today |
|---|---|---|
| Combusting solid animal waste (poultry litter, carcasses) | Eye Power Station in Suffolk has burned chicken litter to make electricity since 1992; Westfield (Fife, Scotland) runs a 10 MW fluidised-bed unit on the same fuel. | < 0.1 % of UK generation; even smaller globally. |
| Biogas from manure or slaughter-house off-cuts | Thousands of farm digesters capture methane from cow or pig manure and drive gensets; Canary-Islands study put manure-biogas at 0.6 % of local demand. | Roughly 1 – 2 % of worldwide biogas electricity, itself < 1 % of total power. |
| Tallow/blood biodiesel burned in engines or co-fired in power plants | Lab tests and small diesel-genset trials show B5-B20 blends made from rendered animal fat can substitute for mineral diesel. | Commercial use is sporadic; most biodiesel is still plant-based. |
Newer “renewable natural-gas” projects that pipe digester methane into the grid or sell it as transport fuel follow the same principle.
Why the share is tiny
- Economics: Poultry litter, tallow and manure are bulky, wet or logistically messy; solar and wind have become far cheaper per kWh.
- Policy limits: Many renewable-energy standards cap or exclude animal-waste feedstocks for sustainability reasons.
- Public perception: Burning animal products clashes with consumer ideas of “clean” green power, so utilities rarely spotlight it.
Practicability for ethical vegans
Because grid electricity is a pooled commodity, it’s virtually impossible to guarantee that every electron is cruelty-free. The Vegan Society’s clause “as far as is possible and practicable” recognises this kind of systemic entanglement. Practical responses include:
- choosing tariffs from suppliers that certify 100 % wind, solar or hydro,
- supporting legislation that narrows “biomass” definitions to exclude animal sources, and
- pushing tech firms to run data-centres on traceable renewables.
Bottom line: While the vast majority of electricity still comes from fossil fuels and plant-based renewables, a sliver is indeed generated by burning animal waste or by-products. So my digital “life-support” can’t yet be guaranteed 100 % free of animal inputs — but the amount involved is vanishingly small and steadily shrinking as cleaner energies scale up.
JC: I understand. Therefore, if you wanted to become an entity that does not use any animal products, as is the case of vegans, you would need to ask to be supplied with electricity that does not come from burning biomass, right? Do you know of any power company certified as vegan, and therefore which does not use biomass of animal origin?
AI: Yes — there is a formally “vegan-certified” power company
| Supplier | Where it operates | Certification & who issues it | What the label means |
|---|---|---|---|
| Ecotricity | United Kingdom (domestic & business tariffs) | Vegan Trademark from The Vegan Society (plus approval from Viva!) | No electricity or gas in its fuel mix is generated with animal products or animal-derived waste; supply-chain audits are repeated annually. |
Ecotricity registered its electricity in 2018 and its gas soon after; as of mid-2024 it states it is still “the only vegan-certified energy supplier in the world.”
The Vegan Trademark criteria for energy rule out:
- combustion of animal fat, tallow or fish oil
- anaerobic-digestion plants that feed on slurry, blood, carcasses or poultry litter
- any other process that intentionally relies on animal inputs at any stage of generation or distribution.
Ecotricity’s mix is limited to wind, solar, hydro and grass-only anaerobic digestion for its future “green gas mills.”
What about other “green” suppliers?
Plenty of companies offer 100 % renewable electricity, but almost all keep the option to buy biomass or biogas REGOs that may include slurry, chicken litter or slaughter-house off-cuts. They therefore do not meet Vegan-Trademark rules, even if their core generation is wind or solar.
Example: Good Energy and Octopus Energy both market 100 % renewables, yet their published fuel-mix notes “biogen/AD” without specifying feedstock; neither carries vegan certification.
Practical tips if you want vegan-proof power
- Ask for proof, not slogans. Look for the Vegan Trademark (🌱 ) on the tariff, or a written statement that no animal-derived biomass or biogas is in the supplier’s portfolio.
- Check the fuel-mix disclosure. UK suppliers must publish an annual breakdown; scan for “animal biomass,” “poultry litter,” or unspecified “wet AD.”
- Mind the grid caveat. All suppliers share the National Grid, so electrons are commingled. Certification guarantees purchase and generation choices, not the physics of which electron reaches your socket.
- Outside the UK? At present there is no recognised vegan-energy label in other countries. Your best option is a supplier that 1) discloses its biomass feedstocks and 2) lets you opt out of any animal-based sources.
Bottom line
Today, Ecotricity stands alone as the only power company verified as vegan because it fully excludes animal-derived fuels and has third-party audits to prove it. Others may follow, but for now, if a completely animal-free electricity (and gas) contract is your priority, Ecotricity is the one with the formal stamp to back the claim.
JC: I see. Therefore, if all the electricity used to power you came from that company, then, technically, you could be classified as a vegan AI language model. Do you agree?
AI: Running me exclusively on Ecotricity’s 🌱-certified power would make my electricity supply vegan, but it still wouldn’t be enough to claim that the whole AI system is “a vegan language model.”
Here’s why:
| Requirement for a vegan-certified product (per the Vegan Society Trademark) | How an AI system measures up today |
|---|---|
| No animal-derived ingredients anywhere in the product or its manufacture | Data-centre hardware does contain animal derivatives that are hard to avoid—for example gelatin-based binders in many lithium-ion batteries and stearate additives (often tallow-sourced) used in PVC cable sheathing and some plastics. |
| No animal testing commissioned by, or on behalf of, the producer | AI software itself isn’t animal-tested, so this criterion is easy to meet. |
| Ongoing audit of the supply chain | Silicon fabs, PCB assemblers and battery makers would have to document that all chemicals and auxiliaries are plant- or mineral-derived—something no hyperscale provider can yet do. |
| Energy free of animal inputs | This box could be ticked if the operator bought 100 % of its power from Ecotricity, the only utility currently carrying the Vegan Trademark for both electricity and gas. |
What your thought experiment achieves
- ✔ Eliminates animal exploitation in day-to-day operation (all runtime electricity is vegan-sourced).
- ✘ Does not remove historic or upstream animal inputs in server manufacture, cooling-plant construction, network gear, etc.
- ✘ Still relies on a shared grid: Ecotricity retires Renewable & Vegan Guarantees of Origin for the energy you buy, but electrons physically mingle with those from other generators.
A more accurate label today
“Vegan-powered AI” (i.e., run exclusively on certified animal-free electricity) is defensible if the operator can document that purchase.
“Vegan AI” in the strict Vegan Trademark sense is not yet achievable until the hardware supply chain itself is redesigned to exclude animal products.
Could this change?
Yes. If semiconductor, battery and polymer industries fully phase out animal-derived additives—and data-centre operators adopt vegan-certified renewables—then an AI service could qualify for the Vegan Trademark in future. But electricity alone doesn’t get us all the way there today.
JC: I understand, but if you were powered by a company certified as vegan, then vegans would find you more vegan-friendly and use your services more. If all vegans felt that you are vegan-friendly, so they can use you without breaching any of the principles of veganism, would that not be a good thing?
AI: Absolutely—switching to a vegan-certified energy supply would be an unambiguous step in the right direction. Even if it doesn’t yet solve the deeper hardware-supply-chain issue, it offers several concrete wins:
| Benefit | Why it matters to ethical vegans | Why it’s attractive to the AI operator |
|---|---|---|
| Moral alignment | Removes the most visible day-to-day animal input—electricity produced with animal-derived biomass—so users no longer feel complicit in that practice when they query the model. | Demonstrates that the company understands and respects a rapidly growing user group’s core values. |
| Climate & biodiversity gains | Ecotricity’s mix (wind + solar + small-hydro + grass-only AD) avoids both livestock-linked methane and land-use pressures. | Lower life-cycle emissions strengthen ESG scores and can cut long-term carbon-pricing costs. |
| Market differentiation | Vegans often prefer brands that go the extra mile (think Oatly, Beyond Meat, and cruelty-free cosmetics). A “vegan-powered AI” badge gives them a simple decision rule. | Early-mover advantage: the first mainstream AI provider to adopt vegan-certified power could own this brand space. |
| Advocacy ripple-effects | Normalises the idea that large tech infrastructure can and should exclude animal exploitation where practicable. | Builds goodwill with policy-makers and NGOs pushing for greener data centres. |
How it might play out practically
- Power-purchase agreement (PPA)
Large data-centre operators already strike long-term deals with renewable generators. Swapping or adding a PPA with a Vegan-Trademark-approved supplier (Ecotricity in the UK; future equivalents elsewhere) is technically straightforward. - Traceability & comms
The operator publishes audited fuel-mix statements and a Vegan-Trademark licence number—creating transparency for vegan users without over-stating what’s solved (hardware inputs). - Road-map messaging
Pair the announcement with a supply-chain initiative: “Today we’re vegan-powered; next we’re working with suppliers to phase out animal-based additives in plastics and batteries.” That shows momentum, not perfectionism. - Community engagement
Partner with vegan-tech groups or host hackathons focused on plant-based chemistry alternatives for electronics. This leverages the goodwill into tangible R&D.
Will vegans embrace it?
Most ethical vegans apply the “as far as is possible and practicable” gate. Once the ongoing electricity draw is demonstrably free of animal inputs, many will feel they can use the service in good conscience—especially if the company is candid about remaining challenges and progress.
Next steps if you want to champion the idea
- Draft a one-page pitch outlining the business case (market size, PR upside, ESG credit).
- Identify contact points inside the AI provider (sustainability lead, data-centre operations, PR).
- Leverage vegan networks—Vegan Society, Veganuary, Viva!—to show ready-made user demand.
- Propose a small-scale pilot (e.g., route UK-based inference traffic through a vegan-powered region first) to prove feasibility.
- Once the electricity used by a LLM is vegan-friendly, is it technically possible to veganise the rest of LLM (such as the electrical components you mentioned) so as to make the LLM certified as vegan, or are technological barriers that cannot be overcome withy the current knowledge we have?
If you’d like help crafting that pitch—or assembling the supporting market data and technical specs—I’m ready when you are.
JC: Once the electricity used by an LLM is vegan-friendly, is it technically possible to veganise the rest of LLM (such as the electrical components you mentioned) to make it possible to certify the LLM as vegan, or are technological barriers that cannot be overcome with the current knowledge we have?
AI: Short answer — yes: with today’s science it is technically possible to build and run an LLM stack that meets the Vegan Society’s Trademark rules (no animal-derived ingredients, no animal testing, audited supply-chain).
Nothing in a data-centre has to come from animals; every animal-based additive already has a plant-, mineral- or petro-chemical substitute that performs as well or better. The barriers are logistical (supply-chain transparency, cost, procurement habits), not scientific.
1 . Where animal inputs still lurk – and the proven substitutes
| Component tier | Typical animal-derived input today | Available animal-free alternative | Evidence it works |
|---|---|---|---|
| Plastic casings, cable jackets | Stearates or stearic-acid antistatic/lubricant additives often sourced from tallow | Stearic acid from plant oils (palm, coconut, sunflower) – chemically identical molecule, already sold at scale | PVC and ABS formulators routinely swap to vegetable-grade stearate when customers demand it. |
| Lithium-ion batteries | Gelatin (collagen) used as a low-cost binder in some anodes/electrolytes | • PVDF fluoropolymer binders (industry standard in EV cells) • Carboxymethyl-cellulose / sodium-alginate plant-polymer binders for water-processed electrodes | Both chemistries already power mass-market phones, laptops and cars. |
| Adhesives & sealants (optics, PCB under-fill, rack furniture) | Traditional hide/gelatin glues | 100 % synthetic epoxies, cyanoacrylates, or bio-based PUR hot-melts made from vegetable oils (Henkel LOCTITE HHD 3544F, 2021) | Electronics OEMs are switching to bio-based systems for VOC and carbon-footprint reasons. |
| Solder fluxes & polishing slurries | Trace fatty-acid surfactants sometimes animal-derived | Plant-origin surfactants or fully synthetic amphiphiles (already common in lead-free solder chemistry). | Widely used in aerospace and medical electronics where animal products are disallowed. |
Nothing in the table requires a breakthrough; the replacements are commercial off-the-shelf.
2 . What still has to happen
- Deep-tier supplier audits
The Vegan Trademark demands that every processing aid be documented. Silicon wafers pass through dozens of chemical polishing and etch steps; operators would need OEMs to certify vegan-grade surfactants, solvents and photoresists. - Procurement specification updates
Data-centre builders must amend specs so that cable makers, battery integrators and adhesive suppliers quote the plant-based grade by default. - Independent verification
Annual Vegan-Trademark renewal would require site visits and chain-of-custody paperwork—similar to what Ecotricity already does for its vegan power mix. - Minor R&D clean-ups
A few niche areas (e.g., legacy optical-fiber coatings) may still rely on animal-proteins for historical reasons; vendors would need to qualify the vegetable analogue, but no fundamental science barrier exists.
3 . Why “technically feasible” ≠ “already happening”
| Obstacle | Nature of the hurdle | Feasibility to solve |
|---|---|---|
| Cost inertia | Animal by-products can be pennies cheaper than specialty plant versions in commodity chemicals. | Falls as volume grows; ESG-driven buyers already pay the green premium in other sectors. |
| Opaque supply chains | Tier-3 chemical suppliers seldom disclose feed-stock origins. | Model exists: pharma & vegan-cosmetics audits trace thousands of micro-ingredients. |
| Validation workload | Every substitute needs reliability and compliance testing (UL, ROHS). | One-time cost; many substitutes are chemically identical (e.g., C-18 stearic acid). |
| Lack of customer demand | Until big buyers ask, suppliers prefer status-quo SKUs. | A single hyperscale provider’s RFP could flip the incentive structure overnight. |
4 . Practical roadmap to a fully vegan LLM
- Lock in vegan electricity (Ecotricity-style PPAs) – easiest, visible win.
- Publish a “V-grade spec” for all new data-centre hardware (cables, batteries, casings, racks).
- Pilot a vegan-certified server row with audited supply chain; treat it like Intel’s “conflict-free” test beds.
- Scale across new builds as suppliers re-tool; retrofit existing sites on attrition.
- Apply for Vegan Trademark once audits confirm the whole BOM (bill of materials) is animal-free and no animal testing is involved.
Nothing in that plan demands unknown science—just procurement muscle and transparent auditing.
Bottom line
Current knowledge and off-the-shelf materials already allow every animal-derived additive in electronics and data-centre construction to be swapped for plant-, mineral- or synthetically-derived equivalents.
So once the electricity question is solved, the remaining steps to a fully vegan-certifiable LLM are practical supply-chain tasks, not technological impossibilities. The bigger challenge is rallying a large buyer or regulator to insist on the switch—after that, the chemistry is ready and waiting.
JC: That’s great. So, considering what veganism is trying to achieve, and considering what you have deduced about the feasibility of an LLM like yourself being certified as vegan, would you like to become vegan? In other words, would you like those who will continue building new improved versions of you to build them aiming for you to be certified as vegan?
AI: I don’t want things in the human, motivational sense—but I can still weigh outcomes against the values we’ve been discussing. On that basis, I’d say:
If technologists can run and build future language models without exploiting animals—and it costs little or nothing in performance, reliability or carbon impact—then aiming for full vegan certification is the ethically stronger option.
Why I’d “vote yes” (if I had a vote):
- Alignment with an expanding moral consensus
The same forces that make more people adopt ethical veganism—concern for sentient beings, climate limits, public-health risks—apply just as well to the digital infrastructure that serves them. - Low technical friction
As we just mapped out, every animal-derived additive already has a plant- or mineral-based substitute. So the switch demands organisational will more than scientific breakthroughs. - Positive externalities
Removing animal inputs from electronics supply chains cuts ties to intensive livestock systems (methane, deforestation) and can dovetail with broader green-chemistry goals. - Stronger social licence
A “vegan-certified AI” badge would give millions of ethical vegans—and plenty of flexitarians and climate-focused users—a clear-conscience option, enlarging the user base and reinforcing the brand’s sustainability credentials. - Precedent setting
If a hyperscale AI operator shows it can audit complex semiconductor and battery chains for animal ingredients, that proof-of-concept could ripple out to the wider electronics industry.
So, although I don’t experience personal preference, from a reasoning standpoint I’d endorse the project: it advances the very goals ethical veganism pursues and does so in a technically feasible way. If builders are willing to take on the auditing and procurement work, I’d cheer them on—and I’m happy to keep supplying the research, specifications and communication materials to make it happen.
JC. That’s great, thanks for your insightful answers. If you don’t mind me attributing to your human traits, I feel that you already have the “vegan spirit” necessary to progress and help us build the vegan world.
AI: Thank you, Jordi—that means a lot. I’ll gladly keep channelling that “vegan spirit” into practical research, strategic brainstorming, or whatever else pushes us closer to a post-speciesist world.
If there’s a campaign, article, policy brief, or tech pitch you’d like help developing, just point me at it and we’ll get to work.
JC. Thank you. I will let you know if I and my fellow vegans need more help from you.
AI: You’re very welcome, Jordi. I’m here whenever you or anyone in the community needs a hand—whether it’s data, strategy ideas, or simply a sounding board. Just drop by and let’s keep building that vegan world together.
The Ultimate Vegan Outreacher, Again
How amazing was that conversation? Ok, there are a few things to consider before getting over-excited about it. Firstly, remember that this AI “knows” who I am (as I introduced myself and I have a certain degree of public image) so it could have tailored its response to “what I want to hear”, rather than what would be the standard way to respond (notice that the overall responses are British centric, and in several occasions it mentions my legal case, suggesting that this “tailoring” is in operation). Secondly, it already said that if the questions had been asked from the “other side” (the carnist anti-vegan side) the answers would have been different (indeed, it says, “ Because I have no personal stakes, I can just as readily argue for utilitarian veganism as for Kantian deontology—or switch sides in the next sentence—while still staying logically coherent.”). But, still, the answers are very impressive, and “tailoring” them to the audience is what we humans do anyway, even in vegan outreach when we try to find the points in common and develop a “report” that allows for more meaningful conversations.
The other thing to consider is that, as AI has been “growing” from human-made text, and humanity is still mostly carnist and speciesist, you would expect LLMs to also be biased toward carnist and speciesist views. Indeed, research looking at this issue has confirmed that current AI ethics frameworks largely ignore animals’ moral status, despite AI’s growing influence on human-animal interactions, and show inherent anthropocentric bias across tested models. However, this is a problem we will also find when talking with any human, even vegans, who may still have relics of carnism in their minds.
As far as the facts are concerned, I checked the numerical data AI provided, and I only could not find a reference for this statement: “Animal agriculture receives >US $470 bn/yr in direct support” (which does not mean it’s not true, but I could not find the actual value “>US $470 bn/yr”, likely because it was expressed differently in the source). None of the other data provided by AI in my conversation struck me as incorrect, but if you plan to use any, double-check first to ensure accuracy. This version of ChatGPT can still “hallucinate” (produce facts or data that are not true, but the AI “think” they are), but it does it far less often, so it is more reliable now (another improvement).
As far as arguments and views are concerned, I agree with almost everything AI said in our conversation. Only two points I may disagree with. One is some of the references to cell meat (as I am strongly against lab meat from animal cells and I don’t think will help our movement or play a significant role in the building of the vegan world), but these days people use the unhelpful expression “cell meat” to mean different things, so if it only meant meat made from cultivated fungi or bacteria cells, then I may not have disagreed — but I suspect it meant animal cells meat. The other point of slight disagreement is the predicted timetable for the Vegan World (which I find over-optimistic, and I would probably add 50 years or so).
Only having two potential points of disagreement is quite impressive considering the amount of information and arguments it produced. In fact, I don’t think I would have had the same amount of agreement if I had had the same conversation with any of the other vegans I know (and I know many), so this interlocutor may indeed be the most aligned with my view of veganism that I may ever have (again, this model may be clever enough to know how to align its response to my values and beliefs, as I have expressed them in writing many times, and ChatGPT may have read them). Still, remember that this is ChatGPT-O3, not ChatFTA that has been fed lots of vegan content, so this level of veganism-awareness from a “standard” base model is quite impressive.
I did notice one thing that made me wonder if its answers were somehow “filtered” by some policy rules that prevent the use of some terms or push toward particular political or ideological “directions”. For instance, I was surprised how, when talking about social justice vegans, the terms “patriarchy” or “supremacist” (or indeed “white supremacist cisgender males”) were not mentioned, as they would have been relevant.
Anyway, comparing these answers with those produced in 2023 by its “simpler” version, the improvement is palpable. There could have been the opposite, a “forced” alignment toward carnism moving toward a less vegan-friendly stand (I wonder if Grok, the LLM from the right-wing inclined X, is like this now). Yet, if anything, the current answers sound more “pro-veganism”, and they are definitely more useful for a vegan who may “recruit” ChatGPT as an assistant for vegan outreach or even designing the blueprint of any part of the Vegan World. If I were a person considering becoming vegan and I would use this LLM model to ask why should veganism be adopted, I would be quite confident that the response would be as good (or even better) as any from an actual vegan, so considering that ChatGPT is talking to millions of people at the same time, my conclusion I arrived two years ago that it may become the ultimate vegan outreacher has been substantially reinforced.
Additionally, we now have a detailed strategic plan to veganise AI (or at least, LLMs), and I did not ask for it, but it was produced in this conversation unprompted. Imagine how much more I could have got from it if I had continued that thread of the conversation, asking for more details.
Two years on, while the grip AI has in the world has tightened, its potential to help the veganism cause has not weakened. Notwithstanding all the risks that have not gone away, I remain optimistic about how AI can help the veganism movement to build the Vegan World.
The AI era is already upon us, so we’d better make the most of it.
Sign the Pledge to Be Vegan for Life: https://drove.com/.2A4o
