Metascience Has the Power to Transform Research — But It Must Serve Society as Well

A major step forward was taken in reshaping the research landscape: the launch of the Metascience Alliance in London. This global initiative unites over 25 institutions—from universities and funding bodies to companies and research organizations—under one purpose: to use metascience to improve the way science is conducted and evaluated.

This new movement gained traction during the Metascience 2025 Conference, which welcomed more than 830 participants from around 65 countries. It reflects a growing awareness that today’s scientific challenges—especially those linked to AI in scientific research, political pressures, and financial constraints—require systemic solutions grounded in evidence and transparency.

Interest in metascience has grown rapidly since the 2010s, particularly as issues related to reproducibility in science and research integrity became more visible. The field now covers a wide scope, including peer review, open science, publication metrics and equity in funding and career development.

A central theme at the conference was the rise of AI in research. From streamlining literature reviews to optimizing grant allocation, scientific AI tools like large language models are revolutionizing how research is performed. But these tools must be studied critically—science AI should enhance, not replace, thoughtful investigation. Metascience offers the framework to assess the impact of AI technologies on research quality, ethics, and societal benefit.

Governments are beginning to act. In the UK, a dedicated Metascience Unit has been established to evaluate and improve the country’s research systems. Still, cultural and institutional resistance remains. Some organizations are hesitant to acknowledge flawed practices due to fears of reputational harm. However, avoiding transparency is more damaging than addressing challenges openly and with methodological rigor.

To maintain public trust, researchers must communicate how science works, not just its conclusions. Clarifying uncertainty, explaining why scientific understanding evolves, and showing how funding serves society are key to rebuilding credibility. Here, open science and metascientific insight can play a crucial role.

Ultimately, metascience must serve more than the scientific community—it must address real-world needs. Optimizing peer review is important, but it shouldn’t overshadow the need for research that solves pressing societal problems. As noted by Tim Errington of the Center for Open Science, the Alliance seeks to bridge these gaps and invite new perspectives.
In a time of global uncertainty, metascientists must rise to the challenge—not only to improve science from within, but to ensure it remains relevant, transparent, and beneficial to all.