The NGI programme
Translated from the French, which remains the reference text. This translation is under review.
“What will the Internet of the future look like? Which building blocks will it need in 2025, to deliver the most economic and societal value? Europe kicks off a new and ambitious initiative on the Next Generation Internet. [...] More than anything else, our future Internet will be free and open, reflecting our values.”
Roberto Viola, Director-General of DG Connect, 26/09/2016
“The European Commission aims to shape this future internet as a powerful, open, data-driven, user-centric, interoperable platform ecosystem.”
NGI, original home page, 14/11/2016
1.1In brief
Next Generation Internet (NGI) is a research and innovation initiative of the European Commission, led by DG Connect, the Directorate-General for Communications Networks, Content and Technology. Its ambition is to rethink and rebuild the architecture and protocols of the Internet in order to make it a "Human-Centric Internet".
Its strategic genesis and the broad public consultations that accompanied it span 2016-2017. The programme was formally launched in 2018, with its first calls for proposals, within the Horizon 2020 framework programme, then renewed under Horizon Europe (2021-2027).
Its method is the direct funding of technology building blocks, mostly Open Source, through a highly unusual and agile grant mechanism, "cascade funding". For its launch phase under Horizon 2020 (2018-2020), the Commission allocated it more than EUR 250 million; under Horizon Europe, tens of millions of euros continue to be invested in it every year.
1.2The context: the quest for a "third way"
The creation of NGI was an act of self-defence in the geopolitical and societal context of the years 2013-2016.
In 2016, Europe acknowledged its strong technological dependence. It was caught between two models of the Internet, both of which it rejects. The American model, that of Silicon Valley, is an oligopoly dominated by Big Tech (GAFAM), driven by a predatory business model based on the mass extraction of personal data, "surveillance capitalism". The Chinese model, that of the BATX, is a state-run, compartmentalised Internet, geared towards social control and censorship. What was at stake for NGI was then to explore a European "Third Way", that is, a sovereign digital space, regarded as a common good and founded on fundamental rights.
To this geopolitical imperative was added the shock wave of Edward Snowden's revelations, in 2013, which traumatised the European institutions. NGI was born of the urgent need to guarantee confidentiality and cybersecurity, in order to restore the trust of European users in their own infrastructures.
The regulatory calendar, finally, is the same. NGI was conceived at the very moment the Union was adopting the GDPR, voted in 2016 and applied in 2018. If the GDPR is the European "legal shield", NGI is conceived as its technological translation: the point is to fund the infrastructure that will make it possible to apply European law "by design".
1.3The genesis
Formulating the political intention and the ideological foundation (2016)
The DG Connect communication entitled "Next Generation Internet - designed for humans and reflecting European social and ethical values" is the original political manifesto of NGI. It releases no funds, but sets out an official diagnosis by the European Commission and fixes an ideological course.
The recognised failure of Web 2.0, first of all. The text records that the modern Internet has strayed from its original promise, that of a decentralised and emancipatory network. The ecosystem has become asymmetric, dominated by closed monopolies, GAFAM without naming them. This business model, based on the extraction of attention and the capture of personal data, breeds a major crisis of trust among citizens: surveillance, algorithmic opacity.
The concept of a "Human-Centric Internet", next. Technology is not neutral: its architecture must be rethought from its design so as to embody European democratic values, transparency, inclusion, pluralism and the inalienable respect of privacy. The individual must no longer be the product, but the master of their digital environment.
Targeting grassroots innovators, finally. The document stresses that this Internet cannot be rebuilt with the traditional institutional actors, the large industrial consortia. It calls for capturing the energy of the "digital mavericks": independent developers, ethical start-ups, isolated researchers and the hacker community of civil society.
Legal and financial engineering (2018-2019)
The Horizon 2020 Work Programme (ICT), through its call for proposals ICT-24-2018-2019 "Next Generation Internet - An Open Internet Initiative", is NGI's first legal and budgetary text. It turns the utopia of 2016 into real credit lines, in the form of a pilot phase within the former R&D framework programme.
Its great technical and administrative breakthrough is "cascade funding" (Financial Support to Third Parties, FSTP), which entrusts intermediary foundations with redistributing most of the funds as micro-grants to the final developers; the section on the implementation strategy describes it.
The call also mandates three technology work strands. It directs research funds (RIA, Research and Innovation Actions) towards precise technical challenges, aimed at breaking the silos: privacy and trust enhancing technologies (PETs), to give the GDPR a technical reality through cryptography and secure messaging; decentralised data governance, that is, distributed architectures, blockchain, peer-to-peer, server federation, to eliminate centralised points of control; discovery and identification technologies, namely alternative search engines without advertising profiling and self-sovereign digital identity systems.
Alongside the code, finally, the text funds Coordination and Support Actions (CSA) to animate this new and highly fragmented community and make sure that these small projects begin to form a genuine NGI ecosystem.
Industrialisation and the imperative of Sovereignty (2021-2027)
The Horizon Europe Work Programme, and more precisely its Cluster 4 (Digital, Industry and Space Work Programme), changes NGI's status. On the strength of the pilot phase, which saw the emergence of tools such as Mastodon or PeerTube, NGI leaves experimentation to become a lasting structural programme, within the European Union's current grand strategic plan.
Attaching NGI to Cluster 4, devoted to industry and space, brings the programme into geopolitics. NGI becomes the linchpin of "Destination 6" (Human-centred and ethical development of digital and industrial technologies). The vocabulary shifts: it is no longer only a matter of making a "fair" Internet, but of building a sovereign infrastructure to reduce Europe's technological dependence. From ethical research, one moves to "strategic autonomy".
The programme also records a fundamental change in its requirements: the consecration of Open Source and of the "digital commons". It mandates the production of open software and hardware (Open Source / Open Hardware). The Commission no longer funds mere projects, it funds the constitution of digital commons (Digital Commons), regarded as a public infrastructural heritage reusable by all, including by the European administrations.
The range of funding widens massively, to cover the whole technology stack (tech stack): the Internet of Things (IoT) and edge computing; open workspaces (NGI Workspaces) and ethical virtual worlds; greening the code, finally, since, in line with the European Green Deal, projects must now prove the energy sobriety of their code and of their software architecture.
The text carries, finally, an injunction to scale up. Drawing the lessons of the hyper-fragmentation generated by the micro-grants of 2018, it demands commercial maturity. Projects must now take part in the international standardisation bodies (IETF, W3C), propose viable business models, and join forces to offer complete alternatives, ready to be adopted by the general public and the market.
The study of these three texts shows that the European Commission proceeded with great method: it first set out the 'Why' in 2016 (rejecting the extractive model by asserting a European ethical vision), then invented the 'How' in 2018 (the innovation of cascade funding, to recruit the vanguard of free software), and finally aimed at industrial Scale in 2021 (Open Source raised to a pillar of the continent's strategic autonomy).
1.4The original vision
The vision of NGI is political, but it accepts that the answer must be technical (the well-known adage "Code is Law").
A vision based on values
NGI rests on a consultation run between November 2016 and January 2017 among 449 people, which made it possible to specify the stakes and the perimeter of the initiative. Eight expectations were put to the participants, each rating their importance from 1 to 5. The Internet should ensure citizens' control over their own data and protect privacy; ensure diversity, pluralism and the right to choose; prevent the concentration of data on a few proprietary platforms; be more resilient, secure and safe, especially for children, the elderly and people with limited digital skills; be more inclusive, participatory and transparent; offer a level playing field for new entrants and new economic models; be a real agent of change towards sustainability; and be more human, finally: social, easy, immersive, emotional.
Each of these expectations received the highest score from at least a quarter of the participants. Three of them stand out: control over data and the protection of privacy, to which 336 participants give the highest importance, diversity, pluralism and the right to choose (295), and the non-concentration of data on a few proprietary platforms (290). The consultation report draws from them a vision statement for NGI:
“The Next Generation Internet should ensure citizen's sovereignty over their own data and protect their privacy as a priority. It should ensure diversity, pluralism and a right to choose and should avoid the concentration of data on a few proprietary platforms.”
Source: NGI initiative - Consultation (2017)
Additional values endorsed, in order of priority:
Participants were also invited to propose additional values. The report classes the statements received into nine types. Two dominate, almost level: protection, safety and trust on one side, inclusion, human rights and Internet as a commodity service on the other. Next comes openness to innovation, then, each carried by a few participants, education, governance, fairness and transparency, auditability and control, and the idea that the Internet should not be mandatory; a few statements, finally, discuss the proposed values themselves.
Schema 1.1 · The additional values proposed by the participants
The report notes that these proposals are dominated by the vision of an inclusive Internet, seen as a basic human right, and one that can be trusted as to its content and as to the protection and safety of its users and their data. Two statements set the tone:
“The internet should be a place for citizens to create and contribute to global society, not merely consume and 'like' content. Citizens should have agency, not 'be the product'.”
“The unfettered free flow of our personal data to just a few siren servers is a distortion. We need a new equilibrium where e-citizens benefit from disclosing personal data to service providers.”
Other participants carry a vision of greater freedoms to innovate on the Internet:
“Internet should allow people to develop innovation in any sectors but also should protect personal data and provide a mean to anyone to withdraw information exhaustively.”
“Internet should be used for public services more widely. Internet should become a critical infrastructure for businesses.”
Between these two visions, the report notes clear tensions, some asking for less regulation and some for less intervention. One statement asks that “the European policy must ensure sovereignty of citizens, enterprises, on their data and digital services to enforce the benefit of all stakeholders in the switching towards the digital society”; another, that “Internet should be out of the hands of governments and their agencies”.
The translation into technical challenges, in order of priority
The consultation finally put seven technology areas to the participants, inviting them to propose two more, and asked them to rate their relative importance. Two areas come clearly first: personal data spaces, rated very important by 347 of the 449 participants, and artificial intelligence (257). Then follow, in a tight group, distributed architectures and decentralised data governance, discovery and identification tools, new forms of interaction and immersive environments, then networking solutions beyond IP; software defined technologies close the list.
Schema 1.2 · The technology areas rated very important
The 2017 report (NGI 2025 Final Report) translates this political ambition into strong technological axes, breaking with the Web as it had developed.
Of personal data spaces (Personal data spaces), first, the report says this:
“In dealing with Personal data we're dealing with the trade-off between the benefits of innovation reliant upon IoT and Big Data and the need to prevent abuse of personal data. The challenge is to keep personal data secure and assure citizens of their control over it whilst implementing infrastructure that enables privacy controls and assigns accountability for its responsible use.”
Artificial intelligence is approached there through its framework more than its technique:
“The technical challenges in AI are easily identifiable from an engineering standpoint, but it is the interdisciplinary approach that will assure a workable ethical framework.”
The five other areas are treated in the same way, as a challenge to meet more than as a technique. Distributed architectures call, according to the report, for interdisciplinary learning: we are accustomed to pyramidal structures with their clear chains of command, whereas these modes of organisation, as Yochai Benkler showed in The Wealth of Networks, can unleash unexpectedly extraordinary results; the hard sciences, mathematics included, and the social sciences will have to collaborate to learn how to grow such networks. Discovery and identification tools are caught, like personal data, in a trade-off between the security of data and the creation of useful services, able to interoperate to provide quality discovery independently of the proprietary platforms. New forms of interaction and immersive environments, which will engage all human senses and abilities, need a human science approach to understand their impacts on society, without neglecting the opportunities they open to European research and industry. Networking solutions beyond IP raise a question of deployment more than of research: the work exists, on the Recursive Internet Architecture (RINA) for instance, and the challenge is to deploy it in the face of vast inertia and vested interests, through innovation in regulation and purchasing policies. Software defined technologies, finally, are related to standardisation and to a strong methodological doctrine that guides NGI participants towards significant results.
1.5The implementation strategy: FSTP, or cascade funding
This is where the NGI initiative is most innovative, and where it departs from the European Union's usual methods.
The 2016-2017 consultations brought a major problem to light: developing a third way for European digital infrastructure, designed as a common good and for its use value rather than its exchange value, meant changing doctrine.
While the Commission traditionally funded consortia made up essentially of large companies and established universities, the focus of these players, combined with the bureaucratic weight of European funding, did not make it possible to direct funds towards the players and projects likely to contribute to an open, sovereign, secure and decentralised infrastructure based on the European values identified above.
Such a project had to be carried out by the commoners and the digital mavericks, the grassroots innovators: independent developers in their garages, ethical hackers, small NGOs, SMEs and tech start-ups. These are players driven and guided by values close to those of the NGI initiative and to "European values"; working in open source, by and for the commons; and, finally, having multiple identities, in small or large companies, associations, the public sector or universities, or not working in the classical sense of the term, a digital commons being thus able to bring together many kinds of expertise, attached to the diversity of the players who contribute to it.
The implementation strategy therefore sought to decentralise the funding itself, through "cascade funding" (Cascade Funding, or Financial Support to Third Parties, FSTP). Instead of managing thousands of complex contracts, the Commission pays large global envelopes to renowned intermediary foundations, such as the Dutch foundation NLnet, with the obligation to redistribute at least 80 % of them directly to the final developers. It is these foundations that redistribute the money, very flexibly, as micro-grants of EUR 5,000 to EUR 50,000. The ceiling was later raised to EUR 150,000 for projects already funded, then recently to EUR 500,000, without this being widely known, owing to NLnet's reluctance to distribute amounts of this kind. The application process is drastically lightened, sometimes a few pages suffice, with answers within a few weeks; the follow-up, on the other hand, is highly qualitative.
Support is centred on the code: NGI essentially funded code, but it also provided practical mentoring, security audits, advice on open source licences, help with accessibility and support in standardising the protocols (with the W3C, the IETF, etc.).
1.6Taking stock of the arrangement (2024)
The most complete assessment of the arrangement is the impact study commissioned by the European Commission and published in June 2024, Benchmarking the impact of the next generation internet initiative - Final study report. It assesses the impact of the first waves of NGI projects against five criteria: impact on society and digital rights, facilitation of EU legislation, impact on standardisation, provision of alternatives to market solutions, and sustainability.
The programme started operationally in 2019, mobilising around EUR 140 million over 5 years to support more than 1,000 projects. The study covers the first wave of projects (2018-2020), extended to the NGI0-Entrust and Core actions, that is 1,014 projects. It rests on two benchmarks. The first, quantitative, is the "NGI 1000" survey, conducted through EU Survey: 239 valid responses, covering 291 projects, a response rate of 28.7% that allows extrapolation. The second, qualitative, "NGI 50", studies about a hundred flagship projects through qualitative data and interviews, to map the technology landscape. We follow its results along four questions: who was funded, what the survey measures, what building blocks the projects provide, what the grantees ask for.
Who was funded
82% of the funded projects are located in the European Union, 10% in associated countries, Switzerland for example, and 8% in the rest of the world. Within the Union, five countries concentrate nearly two thirds of the projects: Germany, France, the Netherlands, Spain and Italy.
Schema 1.3 · NGI projects in the European Union
The projects are spread among the Research and Innovation Actions (RIAs) that funded them. Three of them carry 18% each: NGI Assure (research on blockchain and distributed ledgers), NGI0-Discovery (search and discovery) and NGI0-PET (privacy and trust enhancing technologies); the others, from NGI TRUST to LEDGER, between 7% and 4%. Some calls were highly specialised, LEDGER on distributed ledgers, eSSIF-lab on identity, while NGI0-Discovery and NGI0-PET now cover the whole technology stack.
Schema 1.4 · NGI projects by research and innovation action
By technology category, funding is concentrated on decentralised solutions (26%), software engineering, protocols, interoperability and cryptography (17%), network infrastructure (11%) and services and applications (11%).
Schema 1.5 · NGI projects by technology category
What the survey measures
The survey of the grantees measures the impact against the study's five criteria.
Society and digital rights, first: 100% of the projects surveyed align with at least one European principle; 74% increase the safety and empowerment of individuals; 59% put people at the centre of the digital transformation, and 57% aim to guarantee freedom of choice online.
EU legislation and policies, next: 90% of NGI projects are linked to at least one EU policy, the GDPR, the DSA and the DMA, or the Cyber Resilience Act. The projects offer alternative choices (47%), contribute to the digital commons (44%), support the GDPR (39%), the Cyber Resilience Act (30%), the DSA/DMA (23%) and EU digital identity (15%).
Standardisation: 51% of the projects collaborated with standardisation bodies, mainly the IETF (18%) and the W3C (16%). Contributions include Web protocols (HTML, CSS, WebAuthn, Solid, Fediverse), DNS (DNSSEC), routing, and the crucial standardisation of "software bills of materials" (SBOM), with formats such as Package URL (PURL) and CycloneDX, for the security of software supply chains.
Schema 1.6 · The standardisation organisations the projects collaborated with
Alternatives and freedom of choice: 57% of the projects explicitly provide an alternative to a proprietary solution. These alternatives cover many fields: productivity and collaboration suites, hardware design systems, social networks, messaging, identity and access management, search engines, DevOps, operating systems, security, cloud, maps, VPN, etc.
Sustainability, finally: the potential targets are consumers (70%), developers (69%) and the public sector (69%). 63% of the projects share their code on a public repository and 34% through a distribution platform (app stores, package managers); 29% use both methods at once. 76% of the projects have an active external community, which shows a significant level of reuse; the multiplier effect of open source attracts talent and fosters innovation (an estimated 80,000 people involved outside direct funding).
After the funding, 74% of the projects continue and 26% stop. 58% of the projects are integrated into a wider open source community or a commercial effort; 17% are merged upstream; 16% lead to the creation of a legal entity or a foundation, about 80 new lasting structures created thanks to the programme, whose business models rest mainly on SaaS and managed hosting (47%) and consulting (42%).
Schema 1.7 · What becomes of the projects after the funding
32% of the projects receive new funding, mostly public, of which 68% is further NGI funding to help them mature.
Schema 1.8 · Where the new funding comes from
Eleven technology clusters
The qualitative analysis, a very rich one, details how NGI projects shape the digital landscape by providing the "building blocks" (digital commons) of the Internet, divided into eleven clusters. NGI grants very often served as critical seed funding.
Open hardware, first: NGI supports the democratisation of hardware design, around 10% of the projects. For printed circuit boards (PCB), it funds alternatives to proprietary electronic design automation (EDA) tools, with widely used projects such as KiCad, Ringdove EDA or LibrePCB. For open chips, it supports hardware description languages (SpinalHDL) and the VexRiscV processor, based on the open RISC-V architecture, used in industry and research, and the manufacture of the first Power ISA chip outside IBM (Libre-SOC). For FPGAs, it supports logic synthesis tools such as Yosys, which is becoming a genuine alternative to vendor tools, and affordable platforms (UL3XM/ULX4M, Icestudio).
Networking and transport, next, that is the evolution of the technologies that provide resilient connectivity. In routing, NGI funds SCION, a highly secure routing architecture isolated against failures, adopted by the Swiss banking sector, and the improvement of BGP security (RPKI, Rotonda). In VPN and security, it improves WireGuard, a modern, lightweight VPN integrated into the Linux kernel, and Rosenpass, to make VPNs resistant to attacks by quantum computers. In mesh networks and critical infrastructure, the Irdest project aims at decentralised communication networks, and the RETETRA project reverse-engineered the proprietary protocols of TETRA radios, used by police forces and by industry, discovering and reporting very serious vulnerabilities worldwide. On the transport layer, it supports MPTCP for reliable transmission, for example a seamless transition between WiFi and cellular networks.
The evolution of Internet technologies: for DNS, the GNU Name System (GNS) offers a decentralised naming system that resists censorship and preserves privacy, and tools automate the deployment of DNSSEC; for browsers, the JShelter extension against tracking and the Servo rendering engine, written in the memory-safe language Rust and supported by the Linux Foundation through the SME Igalia; for email, the Lightmeter tool, which simplifies the management of servers.
Web 3.0 and cryptography: for payments, NGI TALER deploys a free digital payment system (GNU Taler) that guarantees the buyer's anonymity while tracing the seller's income to prevent tax evasion, with pilot tests with banks in Germany, Hungary and Switzerland. The study describes its path:
The NGI Taler project has benefited from a progression of funding, starting with small-scale initiatives, followed by upscaling, and ultimately piloting use cases with a substantial budget. The initial funding was small-scale, including NGI0-PET (1st seed), NGI-Assure, and NGI-Entrust, which focused on seed funding and roadmapping activities. This allowed the project to start its development and lay the groundwork. Afterwards, the project received additional funding for upscaling through NGI-Pointer, which provided a larger budget of 150,000 euros. Finally, the Taler project secured a significant budget of over 5 million euros to pilot various use cases.
In the context of the pilot, the implementation of the GNU Taler payment system in Switzerland involves collaboration between Taler System (TSYS), a Swiss bank, a Self-Regulatory Organization (SRO), and the FINMA (Swiss regulatory body). TSYS will integrate GNU Taler with the core banking system of the Swiss bank and become a member of an SRO. This collaborative effort aims to facilitate the successful launch and operation of the GNU Taler payment system in Switzerland, ensuring compliance with regulatory requirements and fostering integration with domestic merchants.
In post-quantum cryptography, NGI funds the integration of quantum-resistant algorithms into OpenSSL (oqs-provider) and the development of KEMTLS and Bertie (verified TLS 1.3).
Web 4.0 and virtual worlds: Wolvic, the only open source browser designed specifically for virtual and extended reality (VR/XR); VR research projects (SearXR), decentralised sharing of 3D models (Manyfold) and immersive 3D navigation (XR Fragments).
Decentralised social media, the Fediverse: NGI provided seed funding to almost all the major projects of this federated network of 3.2 million active users, based on the W3C ActivityPub standard. Flagship projects: Mastodon (microblogging, used by the EU), PeerTube (video), Lemmy (an alternative to Reddit), Kbin, Pleroma, as well as the Spritely project to secure the network and the Scuttlebutt peer-to-peer network.
Instant messaging: funding for the generalisation of end-to-end encryption (E2EE) on the XMPP standard (via OMEMO), for clients such as Conversations, Dino or Monal; the development of components for the future IETF Messaging Layer Security standard (openMLS) and of email-based messaging (DeltaChat, WebXDC).
Collaborative tools: substantial support for CryptPad, an end-to-end encrypted office suite, which the German government notably reuses in its sovereign office suite openDesk; support for the underlying technologies (CRDT, through the Interpeer project) to enable real-time collaboration without a centralised server.
Participatory data and AI: in mapping, massive funding of the OpenStreetMap ecosystem and of consumer applications such as Organic Maps, which has passed one million users, replacing Google Maps; in generative AI, AI Horde, a participatory virtual cluster that lets the general public use AI without owning expensive computer hardware; for legal data, Open Terms Archive, a tool that tracks the terms of use of the GAFAM, used by the European Commission to enforce the DSA.
Digital identity and signatures: massive funding (through eSSIF-Lab and TRUSTCHAIN) for digital identity wallets based on Decentralized Identifiers (DIDs) and Verifiable Credentials (VC), compatible with the blockchain ecosystem; support for the Nitrokey hardware authentication key (an open source FIDO2 key, used by Linux kernel developers), for OpenPGP encryption (Gpg4win), and for decentralised credential systems such as Yivi/IRMA (widely used by the public sector in the Netherlands); quantum protection of passwords (SPHINX, OPAQUE).
Software supply chain security, finally: in reproducibility, projects around NixOS and Guix guarantee that compiled code corresponds exactly to the source code, eliminating software "black boxes" (the GNU Mes project); in software forges, decentralised alternatives to GitHub (Forgejo, Sourcehut, ForgeFed); in composition analysis, the worldwide standardisation of the Package URL (PURL) identifier, which has become the de facto global standard for software bills of materials (SBOM), and the funding of the vulnerability and licence detection tools VulnerableCode and ScanCode, used by most of the large European manufacturers, automotive and telecoms.
What the grantees ask for
The study surveyed the funded developers on how to improve the NGI programme; their answers fall into seven types.
Schema 1.9 · What the grantees ask for
Continue the current approach, first (28%): the grantees overwhelmingly endorse the NGI programme. The administrative lightness of the application process is seen as a major advantage over conventional European grants, allowing small organisations to compete, and the intermediaries (RIAs) play their role well. Help with user adoption and scaling up (21%): projects struggle to compete commercially with proprietary solutions, and the grantees ask NGI to fund "non-coding" roles, marketing, public relations, technology transfer, public education and mentoring. Technology and knowledge (17%): support for basic infrastructure, accessibility (WCAG), zero-knowledge proofs, countering AI-generated content, and securing code with a view to the Cyber Resilience Act; it is also suggested that the EU be represented in the standardisation bodies (IETF). Adapting the funding processes (14%): diversifying the size of grants, from micro-tickets up to grants of as much as EUR 5 million to scale up, easing access to "follow-up funding", simplifying the monitoring, fewer deliverables and more real impact, and offering upfront payments (advances). Maintenance and maturation (9%): a major warning on viability, funding often concentrating on new features; it is essential to fund the "thankless work", bug fixing, documentation and community management. The scope of the funding (9%): the need to fund short-term operational costs, legal fees, compliance, infrastructure, and a request for education programmes and for larger-scale software initiatives. A shared NGI infrastructure (2%), finally: providing free hosting and computing resources (CI) so that projects can offer "demos" at no technical cost to early users.
One respondent sums up the grantees' judgement:
“The NGI initiative has been *brilliant* on so many levels […]. As a civil society organisation, we have won many competitive grants from different funders and programs over the years. But as an example, NLNet's NGI grant scheme is by far one of the leanest, lightest weight application processes we've seen. It combines this process with high expectations as to the creative and technical content, and friendly check ins along the way. As recipients, we can see they've been refining their processes very diligently, carefully paring away time-wasting forms and hurdles, but retaining meaningful questions.
The result is that smaller organisations can actually compete. For so many grants, an organisation literally has to have a full time staff member grant-writer - or a team - just to do compliance and file grant applications. That stifles innovation, and makes people at smaller orgs - technologists and decision-makers - afraid to try new or brave ideas. The light-weight demands of this NGI funding application process wipe those problems away. The funds are not huge, but the program shows if there is good progress in thinking, coding and innovation, there is an open door to apply for funding the next step on the path. Process matters a lot if you want good applicants, and thus quality impact. It's a very smart structure.”
Comment from the survey.
The study concludes that the NGI programme had a significant impact in shaping a sustainable, sovereign technology landscape aligned with the EU; that grassroots projects contributed to protocols, to international standards (W3C, IETF) and to European legislation, by offering alternatives to the existing technology monopolies; and that the multiplier effect of the open source community fosters collaboration and innovation, the grantees' recommendations having to guide the future of the programme: support for maintenance, commercialisation, scaling up.
Our conclusion
When the NGI arrangement was launched, the context was VUCA (Vulnerable, Uncertain, Complex, Ambiguous), and no serious alternative existed: many fragmented, scattered initiatives, useful, more or less used, more or less usable, sometimes necessary, often promising, but always underfunded. One had to pay "to see", and NGI paid, which enabled it to identify the potentials, the most promising projects, and to empower, support and accompany them.
In doing so, it acted as an incubator (presumably the largest in open source worldwide?) and funded a kind of primordial soup, a sum of isolated solutions that does not yet constitute a genuine interoperable and decentralised application ecosystem. The question of the stack, that is, of how the different building blocks funded by NGI fit together, is not yet identified in this report dated 2024. It is at the heart of the future of NGI, the subject of chapter 3.