Building Trustworthy AI: How 28DIGITAL Is Helping Turn Europe’s AI Regulation into Market Reality

Europe’s approach to artificial intelligence has always been distinctive. While other global powers have prioritised scale, speed, or computational dominance, the European Union has insisted on a different competitive advantage: trust.

With the adoption of the AI Act — the world’s first comprehensive regulatory framework for artificial intelligence — Europe has established a new benchmark for how AI should be developed, deployed, and governed. Transparency, accountability, human oversight, and fundamental rights are no longer aspirational principles; they are becoming market requirements.

This shift creates both a regulatory challenge and an innovation opportunity.

Across its growing AI portfolio, 28DIGITAL is helping European startups, researchers, and technology builders translate the principles of the AI Act into practical tools, platforms, and ventures. From certification frameworks and trusted infrastructure to human-centric AI systems and next-generation GenAI entrepreneurship, the KIC is supporting projects that are not simply compliant with Europe’s AI strategy — they are actively shaping how it is implemented.

DEPLOY AI is one of Europe's flagship AI initiatives. Backed by €28 million under the Digital Europe Programme and coordinated by Fraunhofer IAIS with 28 partners from 13 countries, it built the AI-On-Demand platform (AIoDP), a vendor-independent platform for trustworthy, European-made AI technologies — reducing dependency on American providers and targeting SMEs and public sector users. Through its involvement, 28DIGITAL plays a strategic ecosystem-building role, mobilising its pan-European innovation network to ensure the platform reaches beyond research communities into real markets.

Its current components include a Business Navigator of validated European AI companies, a Marketplace for verified AI products, and an Industry Stack with a GenAI Gateway and HPC integration under development. The AIoDP has just launched its first Open Call, offering up to €60 000 to AI Providers to integrate their trustworthy, ethical, and transparent AI products on the platform.

CERTAIN (Certification for Ethical and Regulatory Transparency in Artificial Intelligence) is perhaps the most directly policy-relevant project in this cluster. Launched in January 2025 with 20 consortium partners across ten European countries, CERTAIN focuses on enabling trustworthy and compliant AI systems across the data and AI value chain. It provides guidelines, technical tools, and solutions to support compliance, assess data quality, measure biases, and protect privacy. Through its pilots across sectors such as biometrics, health, energy, human resources, finance and IT, the project contributes to the practical implementation of the AI Act. Within the project, 28DIGITAL leads stakeholder engagement, capacity building, and strategic impact evaluation, supporting adoption and uptake across the ecosystem.

TANGO : Trustworthy AI is not only about compliance; it is also about how humans and machines work together. That is the focus of TANGO, a €7 million project funded under Horizon Europe. TANGO develops theoretical and computational frameworks for AI systems designed to augment — not replace — human judgement in high-stakes environments. Its real-world pilots span surgical decision support, pregnancy care, credit risk evaluation, financial fairness, and public policy design.

By producing open, reusable software frameworks and validating them in complex operational settings, TANGO offers something rare in the AI ecosystem: a scalable blueprint for genuinely human-centric AI systems.

AI-BOOST (Artificial Intelligence for Better Opportunities and Scientific Progress) advances the state of the art through open innovation competitions co-designed with industry stakeholders — structured challenges that combine funding, infrastructure and community-building to accelerate the development and validation of AI solutions. Its Large AI Grand Challenge, developed in collaboration with the European Commission and EuroHPC, puts European supercomputing infrastructure to work on real-world problems: winning teams from the 2024 edition included Lingua Custodia (AI for financial services), Unbabel and Tilde (multilingual AI technologies), and Textgain (AI for analysing unstructured data and harmful content). By pulling researchers and developers from across the ecosystem into a shared challenge framework, AI-BOOST directly addresses one of European AI's most persistent structural weaknesses: fragmentation.

MANOLO complements this by developing AI systems that are not only trustworthy in their outputs but computationally efficient and deployable across cloud-edge architectures. Its pilots are grounded in physical devices and real environments — robotics in manufacturing and healthcare, smartphones in telecommunications, and wearables in further healthcare applications — with results expected in the final quarter of 2026. The full MANOLO framework is scheduled for release in September 2026, after which it will be open for external testing.

LLM-BRIDGE targets one of the most consequential gaps in the European AI landscape: the continent's lag in generative AI. Rather than addressing this at the research or policy level alone, LMM-BRIDGE works directly with entrepreneurs, supporting the full lifecycle of GenAI startup formation — from initial education and mentoring through to growth — alongside the tools and support structures that early-stage founders need. It is a project built on the understanding that closing Europe's GenAI gap requires not just better models, but a stronger generation of builders. Startups working in Generative AI, LLMs, NLP, and related technologies can apply to join the LLM-BRIDGE Venture Incubation Programme and receive up to €25 000 each. Applications are open until 1 June.

EMAI4EU (EMotion Artificial Intelligence specialists for Europe) trains a new generation of European specialists in emotion AI — systems capable of recognising and responding to human emotional states. The ability to build AI that can read, interpret and appropriately react to human affect is increasingly recognised as a foundational competence, with applications ranging from mental health support tools and adaptive learning platforms to more natural and effective human-computer interaction. By investing in the people who will develop these systems, EMAI4EU addresses a skills gap that is as strategically significant as any gap in research infrastructure.

Building an Inclusive AI Ecosystem

Beyond individual projects, 28DIGITAL also contributes to broader ecosystem initiatives, including the EIT AI Community and the Women in AI Thematic Working Group within the European Digital Innovation Hubs network.

These initiatives focus on embedding responsible AI practices across Europe’s innovation landscape while addressing one of the sector’s most persistent structural challenges: gender inequality in AI leadership, entrepreneurship, and employment.

At a time when the AI Act is moving from legislation to implementation, Europe’s competitiveness will depend on more than regulation alone. It will depend on whether startups, researchers, and innovators can build systems that are not only powerful, but trusted.

Discover more