Key points
- AI can accelerate every stage of research and development.
- Its benefits are significant but uneven. And technical and ethical risks can weaken trust.
- Governments need adaptive policy frameworks, collaborative approaches and strengthened governance mechanisms.
- International cooperation is essential to widen access to skills, data and computing resources while safeguarding AI with ethics and accountability.
Artificial intelligence (AI) is reshaping research and development. It can speed up data analysis, help generate hypotheses and improve experimentation. But it also raises concerns about lack of explainability, data privacy and the potential deskilling of researchers.
The report examines how developing countries can use AI to strengthen science, technology and innovation while minimizing risks. It calls for flexible and inclusive policies, supported by responsible AI and data governance.
AI can accelerate research across scientific fields
AI can make research faster, more adaptive and more efficient. One study cited in the report estimates that it could double the pace of research and development and unlock up to $500 billion in value each year. In a Nature survey cited in the report, more than half of responding researchers expected AI tools to become very important or essential to their work.
The applications are already wide-ranging.
- AlphaFold has predicted the structure of around 200 million proteins.
- In materials science, an AI model identified more than 2.2 million new crystals, equivalent to nearly 800 years of traditional research.
- AI systems are also improving weather forecasting, climate modelling and early warnings for extreme events.
Governments should support access to high-quality data, digital infrastructure and skills so researchers and institutions can use these tools effectively.
AI’s benefits come with serious risks
AI’s impact is not evenly shared. In one study, the output of top researchers nearly doubled with AI support, while the bottom third saw little benefit. Researchers with prior familiarity with AI models were 17 times more likely to integrate them into their work.
The risks include biased or poor-quality data, opaque decision-making and difficulty reproducing results. Generative AI can produce fabricated claims, findings and sources.
There is also a growing risk that increased reliance on AI could weaken researchers’ critical thinking and problem-solving capacities. A productivity study found that 82% of scientists surveyed reported lower job satisfaction, citing reduced creativity and underused skills.
It is crucial to keep researchers involved in scientific decisions. Policies should also strengthen transparency and reproducibility and provide clear safeguards for data and privacy.
Innovation policy must adapt faster
The report warns that traditional policymaking can become outdated before new policies are fully implemented. Governments need flexible tools such as regulatory sandboxes, updated intellectual property rules and real-time policy monitoring.
They should also foster collaboration and inclusivity by funding interdisciplinary research, establishing AI-focused research centres and expanding public and academic training. Responsible governance should promote high-quality, diverse data sets, accountability and public participation.
International cooperation can support developing countries in accessing data, digital infrastructure and cross-border research partnerships. Open science and open innovation play a key role in democratizing knowledge and promoting collaboration. The report cites an estimate that about 55% of available AI models had publicly available parameters as of April 2025, making them easier to use and adapt.
Turning AI into a shared development tool
AI can become a driver of inclusive progress only through global cooperation grounded in openness, capacity building and shared ethical principles. These measures can support research that is more open, reliable and responsive to development needs.
```