Anthropic adds Blender Support, Joins the Blender Development Fund as Corporate Patron
Anthropic, the creator of AI tools like Claude and Claude Code just added Blender to their connector directory and started financially supporting Blender. And I don’t know how I feel about this.
There is an update to this story – read it here.
The Blender Lab was already working on an MCP server – a connector that gives AI access to a list of ‘tools’ and lets it use information from your scene or make updates to it:
It offers a natural language interface with Blender’s Python API, improving access to documentation, and allowing users to explore and understand complex setups.
You can see the full list of currently available tools here. Most tools look fine to me, as it looks like it currently can’t return actual 3D data to the AI server, but it CAN take screenshots of your work and transmit them unsupervised and execute Python code on your behalf (see my thoughts on that below).
The Blender Foundation’s Position
It seems the Blender Foundation is taking the position that Anthropic’s support will help develop the Python API (which is clearly a benefit to the AI giant), and freedom:
This support will be dedicated towards Blender core development, to maintain and continuously improve foundational features like the Blender Python API, which enables developers and artists alike to extend and improve the software for custom workflows.
[…]
Blender Foundation’s mission remains to empower artists with free/open source technology and tools. Yet, we also maintain APIs for individuals and corporations to extend Blender, also beyond what’s aligned with Blender’s mission. We consider this part of the Software Freedom that’s embodied with Blender’s GNU GPL license.
This rationale seems a bit thin to me. As the debate on this topic has started to rage already I’m sure we’ll be seeing follow-up comms from Francesco about this.
My View
This support by Claude will make it easier for new Blender users to set up Python scripts and carry out complex operations in Blender quickly. AI can also be a useful tool for specific tasks and can help speed up projects.
However, there’s a sharp tension between AI tools and especially the Artist community, where the tools that have illegally trained on their art are now taking their work. Enabling and even fueling this development might be putting the artists’ future at risk.
I’m sure the Blender Foundation gave this a lot of thought before entering the partnership, and they have other Big Tech partners on board. Now of course a joint PR won’t contain any critical notes, but it would benefit everyone if the Blender Foundation shares their position on the use of AI, what they considers healthy and ethical, and which companies practice ethical data collection for training their models.
As it stands, to me personally “AI” does not align with “Freedom”, or with the general ethics of the Blender community.
Some Final Words of Warning (read this!)
The Blender MCP provides the execute_blender_code tool, and allowing an AI agent to execute code unsupervised is a Bad Idea.
Remember that Blender has OS-level un-sandboxed access (which has always been a weirdly unpatched security hole in Blender, but I digress), and that a rogue agent can wreak havoc on your system. Now you know.
Thank you very much for sharing your thoughts.
Even if your sponsors do not have direct control over the development, I also find it very concerning that the Blender Foundation is now accepting financial support from such companies, when even some commercial companies are officially stating that they will refrain from using A.I. in there products at least until potential ethical and legal issues are resolved in a way that does not harm artists. I honestly would have expected the Blender Foundation to give such decisions more careful consideration…
“AI” does not align with “Freedom”
You are not forced to use it, people have the “FREEDOM” to use it if they wish.
Anthropic’s integration with Blender is an “opt-in assistant” rather than a forced or default part of the Blender interface.
Optional Installation: If you don’t install the MCP add-on, nothing changes in your Blender experience.
You’re twisting my words – my observation is about AI in general. And to follow your analogy: you can’t opt out of the illegal data crawling that these companies do. They train on YOUR art (assuming you posted it online somewhere) and then re-sell its creation as a skill back to you.
Closed source propietary AI and paid tokens are not freedom.
And once again, the ai bros with absolutely no 3D skills are the ones that are going to push Blender beyond its original mission, in inharmonic disruptive ways.
Lets not ignore the fact the AI company that they have chosen to align with has provided tools for a government to create hitlists that included innocent targets during an illegal war.
Wasn’t that OpenAI?
Nope.
https://www.theguardian.com/technology/2026/mar/01/claude-anthropic-iran-strikes-us-military
Oh dear. Thanks for sharing that link.
Thank you very much for the link @DIS-A-RAY, I saw it after posting my reply to Bart.
“In his capacity as Anthropic’s CEO, Amodei often writes on the benefits and risks of advanced AI systems.[5] He is a proponent of an “entente” strategy in which a coalition of democratic nations use advanced AI systems in military applications to achieve a decisive advantage over adversaries while sharing the benefits with cooperating nations”
I don’t know the answer, but that’s is pretty close:
https://www.thatprivacyguy.com/blog/anthropic-spyware/
I saw this as well. The Hacker community can’t decide whether it should be defined as Malware or Spyware, which I found quite funny, because one is obviously better than the other… not. It’s dodgy on so many fronts.
Thanks for the article. I didn’t know it was this bad.
Best options we have are self hosting or some open source AI.
Every draconian shit is first “OPT IN/OPT OUT” up to the consumer, but then it is mandated. This happens every time individuals give up even a little bit of their rights. No crap is force fed to you, at first it is always “your choice” and then you realize that becomes your only “option”.
Trojan horse approach to governing. The majority of society is sleepwalking into a dystopian future that will be very hard to undo. ignorance is bliss.
I think that this is sad beyond limits.
There is already many registers of artists depressed and dying by suicide because of this kind of exploitation.
Not to mention Blender Foundation linking it’s finances to something tied to an speculative bubble.
Dark times.
The ‘Large’ in Large Language Models implies only large corporations with lots of computational resources and storage have the means to train Large Language Models.
That means they become the arbiters of truth of any AI output, even the models running locally instead of on their data centers (because they use a model trained by them, with THEIR weights, which are NOT open).
So far these companies have not shown themselves to be good shepherds of that information either. We found out very early on that not only they ingested tons of copyrighted material, they also ingested tons of disturbing material.
Then there is the humongous climate impact – which is undeniable. You can argue you can run a model locally and it doesn’t use that much power, but that’d be a bad-faith argument of couse, because you can see the data centers springing up left and right to support this new fad, they are not made up. That argument also conveniently ignores the power used to train the model in the first place. It also is not how these things work in practice. A typical query to a commercial chat-bot, is not a single query to a single agent, but instead fires off a query to a network of agents and tools, with output feeding back into itself. They have to fake “intelligence” by having AI check the output of AI (several times), having it talk to itself about your query, having a ‘safety’ ai monitoring the final output, and having a plethora of hardcoded scripts, tools and responses to pick up the edge cases where AI is known to fail (math, etc,…). That is how it actually works in practice.
The societal impact, often dismissed, is important and huge as well. People are already losing their minds over social media – The spread of mis-information there is partially responsible of a worldwide rise of populist governements, and a new concern of rising fascism everywhere. That problematic political distruption is closely tied to the same hoard of new tech-feudalists, who happen to be the same companies driving a lot of the AI tech.
This can only end in disaster.
We can’t just have a defeatist attitude about it and claim ‘ai is inevitable’. It is not. If any technology were automatically inevitable, we’d all live in a nuclear wasteland right now. (Nuclear prolifilation is controlled and limited around the world because it can have devastating effects to humanity too. Admittedly with varying success and sometimes dubious political motives.)
At the end of the day, if you look at the PLETHORA of problems surrounding LLM’s, the only conclusion is that there is not, and cannot be ANY ethical use of them. It’s impossible.
It’s certainly possible you can find them useful, and they help you do things you couldn’t do before. Like creating art, or writing if you’re not an artist or a writer. But maybe try by yourself anyway, that’s how humans are supposed to learn. Maybe reach out to a writer or artist to help you, talk to a human to teach you, instead of a robot. That’s how we’re supposed to interact. If you think that’s too slow and you don’t have time, maybe consider how that’s not a good thing. If you think it puts you at an unfair advantage versus people who do use AI, maybe consider that a lot of us won’t go anywhere near a business using AI. Charge a little bit more for your output, some of us are willing to pay the difference. I’m not saying it won’t be harder, but it’s the right thing to do.
In the end though, Anthrophic is so blatantly evil – and the decision at the blender foundation has clearly already been made, I don’t think they will listen to any of this, beyond maybe making a statement of platitudes to explain away some of the backlash.
Unfortunately the blender foundation isn’t the only F/OSS foundation making this kind of mistake. (Linux foundation? … Torvalds certainly is a tech-feudalist bootlicker now as well it seems) – it’s what happens though. The entire software industry is swinging this way, so it’s full of pro-ai people, and it’s inevitable for large, visible organizations to get infiltrated by pro-ai people.
This is also how regulatory capture works. Lawmaking bodies and governments invite “industry experts” when trying to make regulations, who inevitably always end up being the very people needing to be regulated.
I’m not pulling any of this out of my rear. I’ve worked in the software industry over 20 years. I got out a few years ago when the LLM nonsense started becoming unavoidable because there’s no more ethical employment in the industry left. And I’m very much not alone in this.
Of course, the LLM nonsense does not mark the start of the software industry turning evil. That decline started a long time ago and has been exponentially gaining speed.
Ironically, we’re now quickly getting to a point where all the real software experts are outside the software industry, rather than in it. That’s probably something what someone ought to be paying attention to, because the industry is suffering a terrible brain drain at this point. And it very well might be irreversible.
Not that they care, because they no longer make products for end-users, or even care to have a functioning company at all. It’s all just capital for the sake of capital. Value-extraction, short-term profit, and an oroboros of giants investing in giants to boost stock prices. They don’t need us anymore. It’s all self-sustaining now.
Incredible post, thank you for sharing your thoughts and experience @jns
Agree with everything except for the last bit:
“They don’t need us anymore. It’s all self-sustaining now.”
Unfortunately is not self sustaining, I recommend you reading the “The GenAI Divide STATE OF AI IN BUSINESS 2025” by the MIT’s Nanda project.
(https://fortune.com/2025/08/18/mit-report-95-percent-generative-ai-pilots-at-companies-failing-cfo/).
And the Techno Feudalistic have access to this enormous volume of money by borrowing from the state, also known as “us”. So, in the end they are using OUR money to build that kind of crap, which is a HUGE problem.
Maybe what is missing here is a bold move? Take the bloody money from ANTRE / “Sink”-tropic and apply it on DeepSeek integration with Blender? xD
The guys on the Blender Foundation forgets that the main “feature / tool” of Blender is social innovation, take that out of the equation and the project is dead.
_________
More links on A.I. (Artificial Imbecility):
Lionsgate’s Attempt to Create Movies Using AI Has Crumbled Into Disaster:
https://futurism.com/artificial-intelligence/lionsgate-movies-ai
___________
Environmental impact:
1) https://andthewest.stanford.edu/2025/thirsty-for-power-and-water-ai-crunching-data-centers-sprout-across-the-west/
2) https://techcrunch.com/2025/10/17/your-ai-tools-run-on-fracked-gas-and-bulldozed-texas-land/
___________
Slavery:
1) https://www.foxglove.org.uk/open-letter-to-president-biden-from-tech-workers-in-kenya/
2) https://www.wired.com/story/low-paid-humans-ai-biden-modern-day-slavery/
___________
Post proudly written by a Human :)
You are right, of course, but what I mean by ““They don’t need us anymore. It’s all self-sustaining now.” is that I was pointing out a shift in the incentive structures in corporations in modern western capitalist societies:
In the past, things were simple: a company had to make good products in order to get sales, in order to get more profit, and people could vote with their wallet. If you didn’t like what a company was doing, or if you didn’t like the quality of their output, you could simply just stop buying their stuff. And if enough people did that, the company would have to change course.
It no longer works that way.
Small startups might put in a lot of work and effort to get a useful product to market. Inevitably, they get bought out, and then value-extraction starts. The new owner will start cutting R&D, and raising prices, to extract as much value from the company as possible, until they are ready to sell to the next company who will go ahead and do the same. This type of value-extraction is all about boosting the short-term /perceived/ value of a company (eg: to boost stock price), almost always with disastrous consequences for long-term viability of the company, and of course, disastrous consequences for customers and employees in the long run. Executives and upper management will bail out long before the shit hits the fan, leaving the people doing the actual work holding the proverbial bag.
This pattern has gotten so bad in tech now, that often startups are founded on fake products that aren’t even a real thing, just in the hopes some investor will come in and give them a bag of cash.
Nothing good, nothing useful, is allowed to exist and persist, without being value-extracted and destroyed.
On top of that, when corporations get so large that they have a critical mass of assets, they can simply invest in other large corporations who buy their products, and vice-versa. Causing a stock-boosting feedback-loop. We’ve seen this with nvidia for example, buying stock in ai companies, and those same ai companies buying nvidia stock.
At that scale, they don’t really need a functioning product, or real customers, or even employees, really. The capital feeds itself.
Corporate incentive structures both at large scale, and tiny scale, are completely broken, and it’s causing a massive societal decline.
—
Post also written by a human, with all the typo’s and spelling mistakes included ;)
Lets hope that if Blender crashes Claude doesnt go rogue and deletes or corrupts all the projects. Eheheh
Epic Games seemed like a behemoth, acquiring this & that company, but now they need money. I hope that same happens with these AI companies. Venture capital money is a trap.
To everyone who commented here: Francesco posted an update this afternoon – read more about it here: https://www.blendernation.com/2026/05/01/update-anthropic-joins-the-blender-development-fund/
If blender doesn’t embrace AI, it will die. Even Claude said this to me about it:
Oh this is a good one. Here’s the full saga:
**What happened:** Anthropic joined the Blender Development Fund as a Corporate Patron. The artist community immediately had a collective meltdown. Blender Foundation then walked it back — they’re now accepting the funds as a one-time singular donation instead of a membership, and Anthropic was informed and “supports this decision.”
**Why it’s dumb:** Per Blender’s own funding policy, corporate participation in the Development Fund doesn’t give donors decision-making power over the project’s direction, and it doesn’t imply any alignment between Blender and the donor’s mission or products. So the community panicked about something that was explicitly, already, structurally prevented.
And the kicker — no generative AI functionality is currently available or planned to be integrated in Blender. It was just money going to core development. That’s it.
The real irony for you specifically: Blender Lab was already working on an MCP connector for Blender — the exact thing you use on port 9876. That exists *independently* of Anthropic’s funding. The community torched the ongoing patronage over vibes while the actual AI integration they’re scared of was already happening anyway.
Essentially Blender caved to the loudest, most technophobic segment of their community and rejected *recurring* funding to accept a one-time check — a strictly worse financial outcome — to appease people who didn’t read their own platform’s funding policy. Outstanding move.
Whenever people write “” said to me, I stop reading.
You lost a battle Claude.
And what it’s worst:
It was on May, The First.