DO Impact: When Social Economy Meets AI, the Real Questions Begin

In Brief

What happened: The European DO Impact project held two workshops at Politecnico di Milano on May 4-5, 2026, exploring how social economy organizations can use data and AI while examining what role these organizations should play in AI governance.

Why it matters: The event surfaces a tension rarely addressed in mainstream AI discourse: social economy actors (cooperatives, foundations, social enterprises) operate with different values than commercial tech, yet remain largely excluded from conversations about data governance and AI development.

The bigger picture: As Europe debates AI sovereignty and the EU AI Act implementation, this initiative asks whether organizations built around social purpose might offer alternative models for data sharing and AI deployment that commercial and state actors cannot.

The questions raised in Milan deserve a wider stage. That conversation continues at Human x AI Europe in Vienna on May 19, where the people shaping Europe's AI future will be in the same room. Join them.

The Disagreement Worth Having

Most AI policy debates in Europe oscillate between two poles: state-led regulation versus market-driven innovation. The DO Impact workshops at Politecnico di Milano on May 4-5 introduced a third actor that rarely gets a seat at the table: the social economy.

The framing deserves attention. According to the event description from Politecnico di Milano's School of Management, the two-day program asked complementary questions: How can social economy organizations use data and AI in impact projects? And what role should these organizations play in building the "data society"?

These are not the same question. The first treats social economy actors as consumers of AI tools. The second positions them as potential architects of AI governance. The gap between these framings reveals something important about where the European AI conversation currently sits.

What DO Impact Actually Is

Before examining the debate, the project itself warrants clarification. TIRESIA at Politecnico di Milano describes DO Impact as a European initiative running from 2024-2026, aimed at empowering SMEs and enabling organizations in the Proximity and Social Economy (PSE) sector through digital solutions and data-driven approaches.

The consortium spans nine partners across six countries: Italy (Politecnico di Milano's TIRESIA, Fondazione Piemonte Innova, Torino Social Impact), Spain (Cluster Digital de Catalunya, Taula del Tercer Sector de Catalunya), Latvia (Social Entrepreneurship Association), Lithuania (Social Business Association), Sweden (Coompanion Gothenburg), and the European-level Diesis Network.

The project's stated goals include enhancing sustainability and efficiency, maximizing social value for beneficiaries, and strengthening competitiveness through advanced skills and networks. Standard capacity-building language. What makes it interesting is the underlying thesis: that combining digital technologies with traditional approaches and advanced data management creates innovative pathways to address societal needs.

The Milan Program: Two Days, Two Framings

The workshop structure reported by Fondazione Piemonte Innova reveals the intellectual architecture. Day one (May 4) focused on practical applications: how AI works, what biases emerge, what ethical implications follow. Fabio Ferrari, CEO of Ammagamma, delivered a keynote on "How AI Reasons." Diego Morra from MIT's Senseable City Lab discussed urban data approaches. Elisa Rubegni from Politecnico di Milano examined inequalities in generative AI and stereotyped thinking in children.

Day two (May 5) shifted registers entirely. Mario Calderini from TIRESIA keynoted on the social economy's role in data and AI governance. A roundtable brought together voices from the European Commission's Joint Research Centre (Diletta Di Marco and Sara Thabit from the Digital Sovereignty Unit), Fondazione Compagnia di San Paolo (Fulvio Bersanetti), and practitioners from the field.

The workshop on "Collaborative forms of data sharing for the social economy" points toward the real intellectual territory: whether organizations structured around social purpose might develop different models for data governance than those emerging from commercial or state actors.

The Underlying Tension

Here is where the debate gets interesting. When Torino Social Impact reported on the project's October 2024 kickoff, they noted a central theme: the recognition that social enterprises currently operate with low levels of technological integration, despite the evident potential digital tools could offer.

This observation contains an assumption worth examining. Is low digital integration a problem to be solved, or might it reflect something about how social economy organizations understand their relationship to data and technology? The answer depends on what one believes these organizations are for.

Social economy entities (cooperatives, associations, foundations, social enterprises, mutual organizations) operate under different governance structures and incentive systems than commercial firms. They often prioritize stakeholder value over shareholder returns. They frequently work with vulnerable populations. They tend to have different relationships to the communities they serve.

These characteristics might make them slower to adopt AI tools. They might also make them better positioned to develop AI governance models that prioritize different values than efficiency and scale.

The Practical Dimension

The DO Impact Italy call for participation offers financial support for social economy organizations to attend workshops: up to €1,000 per event, with a total budget of €40,000 for the Italian program. The amounts are modest. The signal is not.

The program targets two categories: SMEs in the social economy (cooperatives, associations, foundations, social enterprises, mutual organizations) and "enabling organizations" (networks, federations, entities that support the ecosystem). The distinction matters. Building capacity among individual organizations is one thing. Building capacity among the organizations that shape the sector is another.

The workshop topics span from practical skills (digital tools, data collection, impact measurement) to governance questions (responsible collaboration in the digital era, building better ecosystems). This combination suggests the organizers understand that technical capacity without governance frameworks produces different outcomes than technical capacity embedded in thoughtful institutional design.

What This Reveals About European AI Discourse

The DO Impact initiative surfaces a gap in how Europe discusses AI governance. The dominant conversation features three actors: states (as regulators), corporations (as developers and deployers), and citizens (as rights-holders and data subjects). Social economy organizations fit awkwardly into this framework.

They are not states, though they often deliver public services. They are not conventional corporations, though they operate in markets. They represent citizens, but through organizational forms rather than individual rights claims. Their governance structures often involve democratic participation by members, workers, or beneficiaries.

The question of whether these organizations should play a distinctive role in AI governance is not settled. One could argue they should simply adopt best practices from commercial and public sectors. One could argue their different governance structures and value orientations position them to develop alternative models. One could argue the distinction matters less than the practical question of whether they have the capacity to engage at all.

The DO Impact workshops do not resolve this debate. They create a space where it can happen.

The Question That Lingers

The roundtable on May 5 included representatives from the European Commission's Joint Research Centre, specifically from the Digital Sovereignty Unit. Their presence suggests the Commission sees social economy actors as relevant to sovereignty discussions, not merely as beneficiaries of digital inclusion programs.

This framing deserves scrutiny. Digital sovereignty in European discourse typically means reducing dependence on non-European technology providers, building domestic capacity, and maintaining regulatory autonomy. Where do social economy organizations fit in this picture?

One possibility: they represent a form of distributed, democratically-governed infrastructure that could complement state and market actors in data governance. Another possibility: they are simply another category of organizations that need to be brought up to speed on digital tools. The difference between these framings has significant implications for how Europe approaches AI governance.

The DO Impact project does not answer this question definitively. It does something more useful: it creates conditions where the question can be asked clearly, with the relevant actors in the room.