Despite work to develop a European Research Area, the EU’s approach to key relevant policies, including AI, R&D, platform regulation, education and learning is deeply fragmented. Even though significant portions of its budget go to support projects in these sectors, legislation continues to be piecemeal and unharmonised. Where the needs of research and (higher) education are considered when crafting new laws, they are often little more than an afterthought. At the heart of this is the fact that the current distribution of legislative competencies does not facilitate law-making focused on research and education.

There needs to be an ongoing focus in the EU’s planning and action on the future. This cannot just be declaratory, but needs structures in place in order to ensure monitoring, accountability, and a constant reflection on how the Union can do better. This is, in particular, the case around copyright, which shapes how information in the internet age can be accessed and shared. We need to ensure that decision-making in this area is focused on public interest goals – not least scientific and societal advancement – rather than almost entirely on the interests of the entertainment industry as is the case now. Internationally, there are strong, working examples of bodies which are part of the executive office, mandated to ensure that law, regulation and policy favours innovation that in turn serves the widest goals of economic and societal advancement.

The Open Data Directive has promoted access to publicly funded data under the principle ‘as open as possible, as closed as necessary’ by default. Despite the widespread take up of open access, much publicly funded research is still off limits to taxpayers, as well as to institutions that struggle to buy the research their own staff have produced. Some open access business models also risk exacerbating concerns about equity and financial sustainability. [7] This creates domestic and international disparities that risk deepening existing economic divides. In the meanwhile, efforts to enable researchers or their institutions to retain rights over their own work and publish them openly are too easily undermined by the imbalance of power that exists between authors, institutions and publishers.

Innovation and competition require not only the right to reuse information, but to access it in the first place. Without this, globally competitive R&D and technological innovation simply cannot take place. And yet, barriers to information access have been created by the switch from copyright law to contracts brought about by the shift from analogue to digital materials. Publishers are routinely refusing to license eBooks to universities. [12] Licensing film and audio is also frequently impossible for educational establishments. Furthermore, researchers do not necessarily find it easy to access data. They are often locked out from the data and algorithms of super-platforms, despite these being vital for their work. Recent efforts to address the issue have fallen short, with the EU’s Digital Services Act, the Data Act and the AI Act offering only very limited and bureaucratic possibilities for researchers. While draft EU health data legislation is better insofar as it focuses on ways to enable access, it is clearly limited only to just one sector.

The reliance on contracts, rather than copyright law, in determining what can be done with digital information is creating insurmountable barriers for research and innovation. In addition to refusing outright to license materials, information providers place strict controls, reinforced by digital locks,[13] on what researchers can do with the materials they hold. As a result actions which are otherwise perfectly lawful for students, researchers and their institutions to perform are frequently removed by technological protection measures and contracts. Even where there is the theoretical possibility to remove or circumvent digital locks, it remains too often unclear in the law how this can be done in reality.

Legislation can be complex and difficult to understand. Even if researchers try their best to interpret the law, they may get it wrong. Faced with the risk of liability, they will often simply try to exclude any risk by limiting the scope of their research or abandoning it altogether. This can seriously harm the public interest – for example when relevant historical sources are disregarded or AI algorithms are not trained on vitally important data because of the fear of litigation. Countries such as the United States and Israel allow researchers to avoid this chilling effect by providing for a limitation of liability on designated categories of organisation such as educational establishments, libraries and archives when acting in good faith.