We like to pretend that science is neutral – that our work is insulated from politics, economics and power. But it isn’t. The funding we receive, the methods we choose and, crucially, the questions we ask – and those we never ask – are shaped by funding, performance metrics, history, social norms, and our own biases. To act as if science floats above these forces is not only naïve, it’s irresponsible.

From the 19th century onwards, scientific institutions were deeply entangled with military power, industrial productivity and colonial rule. Funding priorities were set around national prestige and economic gains. Universities adapted by building reward systems around publications, citations, patents and grants. Generations of researchers were trained to see themselves as specialists accountable primarily to peers and promotion committees not to the communities affected by their work. Within this system, a comforting narrative took hold: “I just do the science; what others do with it is not my responsibility.”

Science that harmed

As a professor, I am painfully aware of the historical lessons that demonstrate how damaging scientific work framed as progress can be.

One example is the many leading scientists and physicians who believed they were helping society while promoting eugenics. In practice, it justified forced sterilisation, institutionalisation, and the brutalisation of disabled people, the impoverished, and racialised communities. Meanwhile, the science was wrapped in the language of public good, making it harder to challenge. The echoes of eugenic assumptions that some lives are less desirable persists today, reminding us of the lasting harm scientists can have on society.

Today, algorithmic systems and social media recommendation systems designed to “maximise engagement” have contributed to polarisation, harassment and mental distress. Built by people who may have wanted to “connect the world” or “democratise access” these systems operate in institutional settings that reward growth, not care.

A pattern emerges: decisions are made far from those most affected and the social, historical, and political dimensions of science are often treated as secondary, if they are acknowledged at all.

Ethical science requires governance that travels with the technology – from research question to deployment – and includes the people and communities most affected.

A manifesto for socially responsible science 

If good intentions are not enough, what would it take to change? A manifesto is, by definition, unfinished – it is an invitation to collective work. But I would like to start the conversation by setting out some clear commitments. 

1. Name the politics of science, openly and early. Make our values explicit in the design, conduct, and communication of research. 

2. Redefine excellence. Treat equity, representation, and long-term social and environmental consequences as core criteria of quality. 

3. Practice convergence, not siloed brilliance. Treat lived experience, social sciences and the arts as essential intelligence, not decorative. 

4. Share power with those most affected. Move beyond token consultation. Communities and groups most affected by research hold real decision-making power over priorities, methods, data governance, and evaluation of impact. 

5. Build systematic processes to surface potential harms, including those that fall outside our immediate field of view.   

6. Align incentives with responsibility. Push institutions and funders to reward collaborative and justice-oriented science – in hiring, promotion and grant allocation. 

7. Refuse institutional silence and require them to align their investments, partnerships, and policies accordingly. 

8. Protect dissent and whistleblowing within science to create a culture that produces socially responsible knowledge. 

9. Educate for responsibility, not just technique. Embed history, ethics, social theory, and critical reflection into scientific training. Students must learn not just how to generate but how to situate it in the world and be transformational leaders. 

10. Hold ourselves accountable. When we invoke “science for humanity,” “precision health,” or “sustainability,” we treat those words as obligations, not slogans. We measure ourselves against the lived experiences of people and ecosystems, not just against citation counts and patents. 

How ethical science is embedded in practice 

At a time when “science” is simultaneously invoked to defend public policy and attacked as elitist, I believe good science needs purpose, accountability and an explicit ethical stance. These do not sit only in labs and universities: they run through funders, regulators, publishers, governments and yes, businesses that decide what gets built, scaled, and sold. 

Initiatives like Imperial’s School of Convergence Science, where I am Co-Director, the Good Science Project, and the “Science for Humanity” strategy, create spaces that prioritise impact and values, question incentives, and redesign how we judge excellence. They will matter insofar as they shift incentives and decision-making: who sits at the table, what counts as excellence, and how harms are identified and acted on.  

The question now is whether we are prepared to turn those intentions into structures, commitments, and convergent practices that make responsibility – not just discovery – the organising principle of what counts as excellent science.