Asia-Pacific AI Investment Hits US$3.8 Trillion Per Firm, Yet 45% Fail to Achieve ROI

Massive artificial intelligence investment in the Asia-Pacific region, averaging US$245 million per organisation over the next 12 months, is creating a risk of resource wastage and loss of competitiveness due to inadequate data foundations and governance. The spending surge surpasses the Americas (US$178 million), EMEA (US$157 million) and the global average (US$186 million), yet organisational readiness is not keeping pace.

The KPMG Global AI Pulse survey of 2,100 senior leaders across 20 countries, released in March 2026, reveals a widening gap between AI investment and outcomes. IDC projects that 45% of AI-based digital use cases in Asia-Pacific and Japan will fail to achieve their ROI targets by the end of 2026, with poor data foundations and unclear value realisation cited as the primary causes, not technological weakness.

The Grant Thornton AI Impact 2026 survey adds another dimension: boards approve investment without setting governance expectations, and leadership teams deploy AI without clear ownership. This has created a ‘proof gap’—a chasm between claims and measurable business impact. More than 75% of organisations in Asia-Pacific and Japan have adopted AI, but few possess impact measurement frameworks beyond activity metrics and pilot project enthusiasm.

For Indonesia, these findings are an early warning. AI investment in banking, manufacturing and retail is expected to rise. However, without orderly data foundations—including data quality, infrastructure and governance—similar failure risks loom. Local startups and companies using global AI APIs from providers such as OpenAI, Anthropic or Google face a double cost burden: token fees in a weakening rupiah and the risk of strategic business data leakage, as recently cautioned by Microsoft CEO Satya Nadella.

Indonesian knowledge workers, particularly in data analytics and technology, will face pressure to upskill rapidly. Conversely, if AI adoption slows due to the execution gap, disappointed firms may resort to hiring freezes or layoffs. The weakening rupiah, trading at IDR 18,060 per US dollar, makes global AI API subscriptions increasingly expensive, while the risk of losing strategic data looms large, especially following Nadella’s warning and Apple’s lawsuit against OpenAI. Indonesian AI startups could either become victims or exploit the gap: developing local models with unique Indonesian context data could provide a competitive edge, but this requires significant investment in data infrastructure that remains limited. Regulators need to formulate policies that protect corporate data while encouraging innovation—a difficult but crucial balance.