The Hon. (Dr.) Harsha de Silva

¶ 01 Hon. Chairman. I was going to speak in the evening, but thank you for giving me the opportunity now.

¶ 02 Hon. Chairman, today we discuss the digital economy under a very important Ministry. While there is much to say, I want to devote my time to one key matter. Before that, let me say this.

¶ 03 We are pleased that, on a decision by the President, experts like Dr. Hans Wijesuriya have been brought in as Presidential Advisors to harness their knowledge to drive this journey. Whatever the political party, digital public infrastructure must be built in this country. Even when we contested the election, a key foundation of our policy was digital public infrastructure. We tried for a long time to build it but, for various reasons, we could not do it to the extent required. If we had earlier implemented a system like India’s Aadhaar, we could have saved public finance leakages and better targeted subsidies. Today, devoid of political division, we look forward to supporting the development of infrastructure for the digital economy.

¶ 04 My question is this. For a long time, many were eager to start an e-NIC. We now have a National Identity Card. My first NIC was lost; I am on my second. The photo shows my hair in my youth. The details can be altered; one could even put someone else’s photo. Security is weak. We all accept the need for an electronic NIC. Successive governments have tried. The concept of the Sri Lanka Unique Digital Identity—SLUDI—was decided years ago.

¶ 05 Consider our Assistant Secretary-General of Parliament, Mr. Hansa Abeyratne. He must be who he is; that must be authenticated—say, when he pays tax at 0 per cent or otherwise. To authenticate, we must collect his biometrics—iris, facial, or fingerprints like in our voting here—and store them in one place. Then, in any transaction, we can instantly authenticate him, like Aadhaar, by building a database of biometrics of all Sri Lankans. Some countries have vendor-specific solutions (Google, Apple), but others have open-source systems. India first built such a system open-source. I went to study it, at the request of the late Minister Mangala Samaraweera, together with then Digital Minister Harin Fernando, ICTA Chairperson Muhunthan Canagey’s successor Ms. Muhunthan?—and others—and reported back. India’s system is not vendor-specific; it is MOSIP—the Modular Open Source Identity Platform. Sri Lanka and India agreed, about three to four years ago, to engage on MOSIP. Through MOSIP, like Aadhaar, we can store citizens’ biometrics. There was a big debate here that if we store biometrics, India will steal our data. Hon. Chairman, as a doctor, you know when we created the “1990 Suwa Seriya” ambulance service after 2015, I was accused of being an Indian agent trying to take Sri Lankan biometrics to India. I repeatedly said it was impossible; they did not accept it then, but they do now. Similarly, some now cry “national security” and imagine spectres. Today we have data of over two million people who used the 1990 ambulance service; not a single record has gone to another country. Under MOSIP, as agreed, Sri Lanka and partners will design and develop a foundational digital ID system with two parts: a biometric-backed ID and a method to authenticate credentials.

¶ 06 Yesterday Dr. Hans Wijesuriya made a presentation to MPs explaining its importance—coupling physical and digital credentials, a unique key, a national data exchange, best-in-class, highly secure, vendor-independent.

¶ 07 Now the issue: while this is good and must be done, in parallel another effort is underway to “digitize” the existing physical ID by issuing a new digital card. Both are being pursued. This is unnecessary. I submit to the House it is a waste of money. When asked whether a digital card is needed, the view was it is not strictly necessary. In India, auto drivers do not carry an Aadhaar card; if needed, they get a printout from a kiosk. It is more a cultural practice. Sri Lanka is different: we use birth certificates and NICs widely. Even then, converting the current NIC into a high-tech digital card with polycarbonate is wasteful if the biometrics and authentication live in the foundational system. Polycarbonate cards cost around US$1–2; PVC cards are about 10 cents. My proposal: if a physical card is culturally desired, issue a simple PVC card. The expensive card adds no value—no chip, no biometrics stored on it—since all that will be in the MOSIP database.

¶ 08 Another point: about 15 million people’s biometrics must be collected. The Department of Registration of Persons (DRP) has already procured equipment, servers and capture devices under the e-NIC process, under the Ministry of Public Security. The foundational ID under the Ministry of Digital Economy using MOSIP is a more complex task—authentication workflows, quality, integrations. Biometrics captured once can be shared with DRP as needed. There is no need for DRP to separately procure equipment and cards to run its own process in parallel. Avoid duplication; reduce costs. We have used this old NIC for decades; we can wait another year and, if necessary for culture, issue a low-tech PVC card. No need for a chip.

¶ 09 Let us think this through. The problem earlier was tussles between two Ministries. If there are national security features, address them separately. But assign this task to the Ministry of Digital Economy, which is building infrastructure to serve security, trade, agriculture, health and many other users. Two parallel projects are unnecessary and wasteful. The total cost is said to be Rs. 20 billion, of which perhaps Rs. 10–15 billion is an Indian grant, and the balance Rs. 10 billion is ours. India’s National Institute for Smart Government is the project manager; MOSIP development is there; a main system integrator can also supply hardware. If we start a separate hardware-heavy project to collect data exclusively elsewhere and then try to cover overlaps here, we risk conflicts in this project. Even if small overruns are tolerable, there is absolutely no need to replicate; it becomes redundant.