On 10 September 2026, Luhut Binsar Pandjaitan, chair of the National Economic Council, said Indonesia's digital social protection programme, Perlinsos Digital, had grown from 43 to 258 districts and cities, and that the President had directed a national expansion. Deputy Minister of Social Affairs Agus Jabo Priyono said the ministry supports using Perlinsos Digital to distribute aid in 2027. Officials say national-scale digital distribution begins in the third week of October, with full implementation of aid distribution in 2027, according to RRI's report. This is one of the largest live tests of digital public infrastructure (DPI) in the region, and it deserves a serious assessment rather than reflexive praise or alarm.
The case for the programme is strong
The strongest argument for it is simple: paper-based welfare targeting fails people in both directions. Ineligible households receive aid, and eligible ones are missed. Officials cite a registration process that once took 75 to 200 days and now takes minutes. More than 4.3 million households have already registered in the pilot districts, per the same RRI report. The ministry also reports over 70% positive feedback from the public.
The design leans on three layers. The first is Digital Population Identity (IKD), a phone-based ID checked against civil-registry data with facial verification. The second is digital payments. The third is interoperable data exchange between agencies. Komdigi, the communications and digital ministry, says the system consolidates fragmented datasets (DTKS, Regsosek and P3KE) and lets citizens self-register through a portal, per Medcom. In Surabaya, the report cites 99.67% IKD activation among citizens. Social Affairs Minister Saifullah Yusuf has said the model removes the need for collective account opening and allows registration with only a national ID and facial recognition, reported by Katadata.
The fiscal claim is large. Luhut has said digitalisation could save as much as IDR 260 trillion (about US$15 billion), a figure relayed by GovInsider. We could not find a published methodology for that number, so treat it as an official projection and not a measured result.
Where the risk concentrates
A pro-innovation stance does not mean waving DPI through. Three risks matter most.
Biometric exclusion. Facial verification against registry photos fails more often for some people, such as older users, manual labourers and people with changed appearance. If a failed match blocks a family from food or cash support, the harm is concrete. The programme's stated goal is to cut exclusion errors, so it should publish its false-rejection rate and give every applicant a human fallback.
The connectivity gap. GovInsider notes internet access of only 9.3% in Indonesia's frontier, outermost and underdeveloped (3T) areas, against over 35% in other rural districts. A phone-first system can quietly shift the burden onto the people who are hardest to reach. Agent-assisted registration and offline-capable verification are not nice-to-haves; they decide whether the programme works where need is greatest.
Data concentration and purpose creep. Linking identity, payments and agency data makes a powerful welfare tool, and it also builds a rich profile of the poorest citizens. Indonesia has a legal base for protecting that data: Law No. 27 of 2022 on Personal Data Protection, which sets out data-subject rights and controller duties. But a law on paper is not enforcement. Its value here depends on an operational data protection authority, clear limits on what welfare data may be reused for, and audit trails on cross-agency queries. The Electronic Frontier Foundation, in a September 2026 essay on digital sovereignty, argues that state control over data and infrastructure must be paired with protections for privacy and encryption against government as well as corporate threats. That principle applies directly to welfare databases.
What proportionate regulation looks like
The wrong response is to slow a programme that reaches millions of households sooner. The right one is to attach enforceable conditions to the expansion:
- Publish performance data. Facial-match failure rates, appeal outcomes and time-to-payment, broken out by district type, should be public.
- Guarantee a non-biometric route. Any applicant who fails verification should have a human-assisted alternative within a fixed number of days.
- Limit purpose. Welfare identity and payment data should not be reusable for unrelated enforcement without a legal process.
- Independent oversight. Audit access logs for the interoperable exchange layer, and publish summaries.
Indonesia is also positioning itself as a DPI leader. GovInsider describes the rollout as a dress rehearsal before the country hosts the Global DPI Summit in March 2027. That raises the stakes, because other governments will copy the architecture and the safeguards, or the lack of them.
The bottom line
Perlinsos Digital is a credible answer to a real problem, and the minutes-versus-months gain is exactly the kind of benefit that justifies public investment in DPI. Nothing in the available reporting shows it is unsafe. What is missing is evidence on error rates and data governance at the scale of 258 districts, let alone the 541 planned nationally. Expansion should proceed, with transparency obligations attached so that speed does not come at the expense of the people the system exists to serve.