Civil-society infrastructure against AI-enabled power concentration, built around autonomous weapons
Civil-society infrastructure against AI-enabled power concentration, built around autonomous weapons
Project Details
Updated 07/10/26 · Edited by orgThis project is about training other organisations around Europe to campaign against autonomous weapons. We run a method we call EduCampaigning: we teach people the argument, we run workshops where they turn it into their own campaign, and then they go and campaign. This grant pays us to build the reusable core of that for autonomous weapons and hand it to civil-society groups across Europe. All learning materials created in this project will be Open Source.
Concretely, three things:
First, a training package: a curriculum, a workshop kit, and a clear briefing on why autonomous weapons are the sharp end of AI power concentration and where the negotiations for regulation stand at the moment, and what can we do to pressure our governments towards regulation.
Second, a plain-language toolkit, messaging and visuals, that explains the "loyalty bottleneck" idea to people who don't read AI-safety papers.
Third, workshops with partner organisations to test the materials and train the first cohort.
This project builds the reusable training core, the open curriculum, the loyalty-bottleneck toolkit, and the first workshops, that the full $400,000 campaign we have proposed to Longview Philanthropy (civil society engagement, public art tour, press relations, and advocacy at the G7, EU Council, and UN) would then scale across Europe; if Longview comes through this grant compliments it magnificently, and if it doesn't, this piece still stands on its own.
Theory of Impact
Updated 07/17/26 · By grantmaking.aiThe version of AI catastrophe we work on is lock-in: a small group getting permanent control that no one can later undo. Autonomous weapons are where that stops being abstract. Right now, if someone wants to hold power by force, they need a lot of people to go along with it: soldiers, police, officials, and any of them can refuse. That refusal is one of the last real checks we have. Once the killing can be done by machines that don't refuse, the check is gone, and a very small number of people can hold power without anyone's consent. We call that the loyalty bottleneck, and removing it is one of the clearest paths from "powerful AI" to "power that can't be reversed."
You can't fix that with a better-behaved model alone. It needs rules, and rules need public pressure behind them, in the specific countries that can push for them. The political will already exists; what's missing is the sustained pressure that connects it to binding action. That is the gap we work in.
This grant sits upstream of all of it. Instead of running one campaign, we build the thing that lets many groups run their own: shared materials, a tested workshop model, and framing an ordinary person understands. Money spent here keeps working after it is gone, because the assets are open and the trained organisations don't need us for their day-to-day execution of campaigning. It is a cheap way to grow the civil-society side of AI Safety, which is thin next to the technical side, and to grow it around the one AI risk that is already loading live ammunition.
People
Updated 07/10/26 · Edited by orgProject Lead
Funding Details
- Jul 1, 2026
- -
- 6 months
- -
- -
- -
- -
- -
- Seeking first grant
- Anti Entropy
Track Record
Via Lattea is a small non-profit in Rome, Italy, working at the intersection of activism, art, and education, operating from three dedicated spaces in Testaccio. It was founded as the campaigning continuation of Latte Creative, a strategic communications agency with 15+ years of advocacy work for organisations including Greenpeace International, Amnesty International, Save the Children, Doctors Without Borders, Oxfam, and the Council of Europe. Its proprietary EduCampaigning model (Education → Workshops → Public Campaigns → Social Change) is an established method already run across campaigns on AI safety, the climate emergency, and human rights. Via Lattea is a member of the Stop Killer Robots coalition (260+ organisations, 70+ countries).
I endorse this approach, at least at a minimum level.
I think that lethal autonomous weapons is a visible and attention grabbing entry point to discussing AI concentration of power risks. I also think that regulation of lethal autonomous weapons is an important stand-alone issue. I'm not sure that lethal autonomous weapons is the most important aspect of concentration of power risk (eg control of AI, surveillance / propaganda / information control, economic concentration etc would all seem more immediate risks). But at best this is an entry point to discuss those risks, and at worse lethal autonomous weapons is important by itself.
The campaign slogan is interesting 'Preserving the Human Veto' at a time when, as you'd be aware, there is significant backing away from the idea of 'human in the loop' as a guardrail.
Hey Mark,
Thank you for your thoughtful comment. I fully agree that autonomous weapons are not the most important aspect of the risk of the concentration of power. I am a communications professional and this project (and the more extended one we applied for 400k from Longview) is built on the idea that we can use autonomous weapons as a type of a Trojan Horse to bring people into the discussion around power concentration and AI Safety in general. At the moment when engaging with the public and students (I teach a yearly course on communications at the Aalto University in Finland), it seems to me that the smartest people who are not involved with AI are "anti-AI" and not interested in engaging with these topics at all. So, a bridge is needed to cross this gap.
At Via Lattea, we are part of Stop Killer Robots coalition, which I have known about for a decade through my artistic practice in relation to AI risks in the military industry. Yet, when Via Lattea became a member and we started to talk about their work, we noticed almost no-one knows about them! This is another gap we wish to fill: communications about the advocacy work in this field and what are the possible regulatory frameworks that can make the world safer.
Have a great day Mark.
Kind regards,
Jenna
Thanks for this additional background Jenna - good luck with your application! And yes, I fully agree with the approach of using lethal autonomous weapons discussions as the attractive entry point / trojan horse.
Hi Jenna, I appreciate your earlier clarifications and I'm happy to endorse the funding of your minimal viable product of $20,000. I see great value in building on some of the earlier campaigns (and lessons learned) and applying environmental, health, human rights campaign approaches to AI safety advocacy and outreach.
Can you please confirm:
Grateful also for your commitment to make the assets publicly available after the pilot workshop.
@Mark Aiken Excited for this grant recommendation!
@Mark Aiken Hey Mark! Thank you so much for your message. So exciting!
Regarding your questions:
\
Assets will be made public after the pilot workshop.