brand logo
Brain farms before data centres

Brain farms before data centres

15 Nov 2026 | By Harindra B. Dassanayake


  • Defunding research and exporting brains
  • Before the data centres came the brain farms. 


Long before the world began building warehouses full of servers for supporting Artificial Intelligence (AI), it was harvesting minds, and Sri Lanka has been one of its most diligent farms. 

The harvest has been impressive. In 2003, Malik Peiris, who trained in medicine in Sri Lanka, led the Hong Kong team that identified the coronavirus behind SARS. Cyril Ponnamperuma led the National Aeronautics and Space Administration’s (NASA) work on the chemical origins of life. Sarath Gunapala, a Colombo University graduate, built infrared detectors at NASA’s Jet Propulsion Laboratory.

Some names have passed into folklore. The Founder of Amazon tells how, as a Princeton student, he spent three hours stuck on a physics problem, then took it to Yasantha, a Sri Lankan classmate he calls “the smartest guy at Princeton”. 

Yasantha glanced at it and gave the answer in one word. The future billionaire gave up physics for business. Yasantha went on to file some 94 patent applications in the United States; at home, the National Intellectual Property Office (NIPO) granted 40 patents to Sri Lankans in the whole of 2024.


The farm, never the field


This is how Sri Lankan names in science are celebrated abroad. Why do they flourish only once they leave? 

Part of the answer is what we spend to keep them. The National Science Foundation’s (NSF) ‘Statistical Handbook on Research and Development’ puts Sri Lanka’s research spending in 2024 at Rs. 30.7 billion, just 0.10% of Gross Domestic Product (GDP), down from 0.16% in 2010. We are the farm, never the field.

Nor is this simply ‘brain drain’. A generation ago, mostly young graduates left. Today it is academics and professionals in mid-career. The Federation of University Teachers’ Associations says that 20–30 lecturers leave every month, often the most experienced. Of the 1,085 doctors sent abroad for specialist training between 2022 and 2024, one in five did not come back.

The State reaches for the bond. On 4 November, the Committee on Public Enterprises (COPE) heard that university teachers who broke their bonds owe nearly Rs. 30 billion. A bond can recover a debt; it cannot make staying worthwhile when policy makes leaving the rational choice.

We pay for a scientist’s education from Grade 1, often fund a PhD abroad, then offer an ill-equipped laboratory, a salary taxed at up to 36%, and a patent office that takes years. We do the farming; the West does the harvesting.

Governments answer with buildings. But research is done by people, and it is the people we are failing to keep.


One rupee in a thousand for new knowledge?


Generations of schoolchildren learnt Kumaratunga Munidasa’s line, ‘aluth aluth dæ nothanana jathiya lowa nonagee’: a nation that does not create new things will not rise in the world. That nation now spends one rupee in every thousand on new knowledge, less than any of its neighbours, and the lowest in South Asia. India spends about 0.6% of GDP, Bangladesh and Vietnam about 0.4%, Pakistan 0.2%, Malaysia 1%, and Thailand 1.2%. 

The NSF’s report presents a measure of the erosion. The Government funded less than a third of that Rs. 30.7 billion; business and foreign sources paid for two-thirds. Sri Lanka has 122 full-time researchers per million people, against 347 for South Asia and nearly 1,500 worldwide.

Despite the rhetoric, budgets have done little to change this. They reorganise institutions rather than invest in people: a national council on research and development, promised for 2026, is now promised again for 2027.

Budget 2027 allocates Rs. 23.8 billion for research, up from about Rs. 21 billion. But an allocation is a promise, not a payment: the Government spent only about half its capital budget in 2025, and just 17.4% by the third quarter of 2026. When last counted, for 2024, the State actually put Rs. 8.8 billion into research.

Even on paper, it is a tenth of the Rs. 251 billion set aside for highways and railways. University staff get a Rs. 25,000 rise in their monthly allowance, paid in two halves; scientists in State research institutes get nothing comparable. The NIPO is not mentioned at all. Expressways have a constituency; laboratories, and the people who work in them, do not. We are a nation that systematically defunds research and says ‘no’ to new knowledge. 


A patent office without examiners


The National Intellectual Property Office has been starved for years. The Auditor General found nine of its 10 patent examiner posts vacant for more than six years. The wait time for a patent can be over two years, which often leads to someone else patenting it elsewhere.

The cost of such delays is counted in lives. Snakebite cases reported in Sri Lanka’s hospitals are around 30,000 and antivenom costs around Rs. 10,000 a vial. In his keynote oration titled ‘The theory and Practice of Corruption in Sri Lanka’ at the Sri Lanka Association for the Advancement of Science in November 2024, Rohan Pethiyagoda described a Sri Lankan scientist’s Rs. 200 blood test that quickly shows whether a victim has been envenomed. Trialled at the Monaragala Hospital, it beat the expensive test in use.

Then, according to Pethiyagoda, the invention was fraudulently patented in India by a Chennai company with no medical research background, leaving Sri Lanka facing royalties on its own discovery. The NIPO, after years, had still not granted the scientist a patent at home. He left for a professorship in Australia.

Pethiyagoda presented it as a case study in corruption, rooted in an absence of trust between people and institutions. When ownership of a public discovery is unclear and the patent office cannot act, those who move fastest, not those who did the work, end up holding the rights. An underfunded patent office without oversight cannot protect inventors; instead, it exports them.


What America learnt in 1980


The United States faced this problem four decades ago. By 1978 its Government held some 28,000 patents from research it had funded and had licensed fewer than 4% of them. 

The Bayh-Dole Act of 1980 let universities keep title to such inventions if they commercialised them and required them to share the royalties with the inventors: the researchers who did the work in the laboratory. It gave academics a direct stake in research that solves real-world problems. 

In 2002, The Economist magazine called it “possibly the most inspired piece of legislation to be enacted in America over the past half-century”.


What should be done


The answer begins with pay. Salaries, not buildings, decide whether a scientist stays. Universities and State research institutes need a research cadre whose pay is benchmarked against the region rather than the general public service, with the research allowance freed from tax and increases paid in full, not spread over two years.

Money must also follow a plan. Sri Lanka should name a handful of long-term research programmes where it has a natural advantage, such as tropical agriculture and medicine, snakebite, marine resources, minerals, and AI for Sinhala and Tamil, and fund them for a decade, beyond the reach of the annual budget cycle and with guaranteed access to national computing capacity. 

The Government already has the promise it needs. Six percent of GDP for education was the rallying cry of the 2012 university teachers’ strike, and the National People’s Power (NPP) manifesto pledges to reach it. The 2025 Budget gave education 1.86%. A credible path to 6%, with a protected share for research rising to 0.5% of GDP by 2030, would turn that pledge into policy.

Finally, researchers must own what they make. The Research and Development Bill should carry a Sri Lankan Bayh-Dole clause that guarantees inventors a share of the royalties. That needs a patent office with examiners, a budget, and a backlog it publishes every quarter, and a State that treats the theft of publicly funded discoveries as the crime it is, prosecuting those responsible and protecting those who report it.

Budget 2027 offers electricity tariff concessions to investors who build data centres and nothing comparable to keep a virologist or a physicist. Servers can be bought or rented; a generation of scientists cannot. Invest in the brains, or we will go on running the farm while others keep the harvest.


(The writer is a policy analyst affiliated with Muragala | Centre for Progressive Politics and Policy [CPPP] and the Lanka Data Foundation)


(The views and opinions expressed in this article are those of the writer and do not necessarily reflect the official position of this publication)




More News..