Vikram Sinha wants more from artificial intelligence than chatbots that speak English with an American accent. The CEO of Indosat Ooredoo Hutchison sees the technology as a tool to solve problems rooted in Indonesian soil. And he intends his company to deliver it.
Sinha has steered the merged operator toward what he calls an AI-native future. The strategy rests on three parts. First, bake intelligence into the network itself. Second, build cloud and security offerings that rely on local data centers. Third, and most visibly, create Sahabat AI. The name means close friend in Bahasa Indonesia. It functions as both a consumer app and an open platform for developers.
The initiative matters in a nation of more than 270 million people spread across thousands of islands. Many speak regional languages alongside Bahasa. Models trained primarily on English or Mandarin data miss nuances. They carry biases that do not fit local realities. “What gets solved in the U.S. or China might not work in Indonesia,” Sinha told Fortune. “The LLM is not neutral. And if it is not in your language, it will have bias, cultural nuances and all.”
Indosat launched Sahabat AI in partnership with GoTo Group. The model now supports multiple languages, including Javanese, Sundanese, Balinese and Batak. A 70-billion parameter version powers it. Users access writing help, research, coding assistance and image generation through Android and iOS apps. The platform also serves enterprises. Banks and government agencies gain tools tuned for local regulatory language and cultural context.
But business logic trails the vision. When Sinha asks his team for a clear path to profit from Sahabat, the answer remains uncertain. “If I ask my team whether they can make a business case for Sahabat? They don’t know how,” he admitted in the same Fortune interview. Doubts persist. Early adoption has not produced the daily or monthly active user numbers that typically drive valuation.
Still, Sinha pushes forward. He believes the payoff will come. Telcos hold unique advantages here. They own the pipes. They sit closest to millions of end users. Inferencing that happens near the edge reduces latency. Applications built for Indonesian problems outperform imported ones. “Telcos like us can take intelligence to the edge with low latency, and then develop applications which are made in a country, for the country, instead of just buying it from China or the U.S.,” he explained.
Partnerships accelerate the effort. Indosat works with NVIDIA on GPU infrastructure and the AI factory. Nokia contributes to intelligent networks. A recent tie-up with Google brings Gemini capabilities to customers through data bundles. That combination helped lift average revenue per user in the first quarter of 2026 to its highest level since the 2022 merger of Indosat and Hutchison 3 Indonesia.
Financial results show progress. Revenue for full-year 2025 reached 56.5 trillion rupiah, a modest 1.1 percent increase. Profit grew faster at 12.2 percent to 5.5 trillion rupiah. The first quarter of 2026 delivered 12.1 percent revenue growth. Core business still depends on connectivity. AI services remain a smaller slice. Yet Sinha argues the combination of connectivity plus compute positions the company for double-digit growth again.
Indonesia offers structural strengths for AI infrastructure. Power, land and water matter when training and running large models. Sinha noted his company holds approvals for nearly 800 megawatts. “The U.S. doesn’t have power,” he said. Such resources give the archipelago an edge over saturated markets in the West.
The government backs the push. Collaboration with the Ministry of Communication and Digital Affairs framed Sahabat AI as part of national digital sovereignty goals. The February 2026 app launch marked a shift from concept to scaled deployment. Officials see it as proof that Indonesia can develop technology relevant to its own people rather than import solutions.
Sinha draws on experience from other emerging markets. He led operations in India, Seychelles and Myanmar before Indonesia. In Myanmar, skeptics warned of talent shortages. He found competence when given investment and opportunity. The same principle applies now. “Without human capital, you will never become sovereign,” he said. “Sovereignty is not only about investment or money.”
The merger that created Indosat Ooredoo Hutchison tested his philosophy. Most telecom combinations fail. Sinha set a different standard. He focused on maximization rather than cost cutting. “The number one guiding principle we wanted to follow was to look at the merger from a maximize, not optimize, outlook: How could we make one plus one equal 11?” Employees and customers responded to growth targets and better service, not synergy spreadsheets.
Broader market conditions complicate the story. Indonesian tech stocks have suffered. Indosat shares fell about 9 percent year to date through early May 2026. The Jakarta Composite Index dropped further. Many observers blame a hangover from the unicorn era. Startups chased valuations instead of sustainable models. “The problem is there were a lot of unicorn startups,” Sinha observed. “They were chasing the wrong metrics. They were all on the valuation game. That mindset has to change. We have to build businesses around more sustainable models with things that are more real.”
Sahabat AI aims to foster exactly that shift. It operates as an open platform where Indonesian developers and startups can build applications. The hope is that a new generation of companies emerges around locally relevant AI tools. Agriculture apps that understand regional dialects. Healthcare assistants familiar with local symptoms and remedies. Education platforms that reflect Indonesian history and values. Success here would create an entire layer of economic activity beyond the telco balance sheet.
Analysts watch the ARPU impact closely. Early signs from AI-enhanced services, including anti-spam tools and bundled Gemini access, appear positive. Yet turning a sovereign large language model into predictable revenue requires patience. Sinha acknowledges the early doubts. “We have not promoted this in a manner where we are driving daily and monthly users,” he said. “We are very confident that a business case will emerge. But yes, there will be doubts in the early days. You have to go all in and make sure you believe in it.”
Recent coverage reinforces the ambition. At Mobile World Congress in Barcelona, Sinha positioned Indonesia as a potential AI hub for Southeast Asia. Partnerships with NVIDIA and Nokia demonstrated AI-native networks, including 5G calls enhanced by intelligence. TelecomTV reported his view that real applications would appear within 12 months using Sahabat as the core model.
A McKinsey interview from late 2025 captured the longer vision. Sinha described Indosat’s goal to help Indonesia become a digital creator rather than a consumer of technology. The operator would move from connectivity provider to enabler of national innovation. That conversation, available at McKinsey, outlined the same three-pillar approach still guiding strategy today.
Challenges remain. Talent acquisition continues. Competition from global hyperscalers looms. Monetization questions linger. Yet the bet feels calculated. Telecom operators in mature markets struggle with flat revenue. In Indonesia, the combination of young population, digital adoption and government support creates different math. AI offers a path to higher margins and new revenue streams.
Sinha shows no sign of hedging. The company invests in sovereign cloud. It deploys GPUs for enterprise clients in banking and resources sectors. Network intelligence improves customer experience and operational efficiency. Sahabat AI serves as the public face and the developer magnet.
Whether the business case materializes quickly enough will determine if this becomes a template for other operators in the Global South or a cautionary tale about overreaching ambition. For now, Indosat moves with conviction. The infrastructure builds. The model improves. Local developers test ideas. And Sinha keeps repeating the core belief. AI should belong to everyone. Especially those whose languages and lives the dominant models still fail to understand.
WebProNews is an iEntry Publication