Singapore leads global per capita use of Anthropic’s Claude AI as businesses ramp up adoption: GIC
SINGAPORE ranks as the largest user of Anthropic’s artificial intelligence (AI) model Claude on a per capita basis, highlighting the Republic’s rapid adoption of AI across businesses, Dominic Soon, senior vice-president in private equity at GIC, said on Thursday (Apr 23).
Delivering the welcome address at a closed-door event co-hosted by GIC and Anthropic, he also said that organisations large and small can benefit “immensely” from AI if it is implemented responsibly, safely and thoughtfully.
In the audience were 150 business and technology leaders with Singapore-based operations; among them were executives and founders of top Singapore startups, investors from Singapore and Asia venture capital firms, Singapore-based corporates and agencies.
In February, Singapore’s sovereign wealth fund led a US$30 billion Series G funding round in Anthropic, alongside US investment firm Coatue Management, valuing the Claude developer at US$380 billion.
Financial Times reported then that GIC was expected to invest about US$1.5 billion in the round, but the final allocation was not publicly confirmed.
AI supporting enterprises
Katelyn Lesse, Anthropic’s head of engineering who spoke at the event, said that as AI capabilities surge, there is a widening chasm between what the technology can do and what businesses are able to implement, and companies are racing to keep pace.
Anthropic executives said model capabilities were advancing far more quickly than business capabilities, creating what they termed “unrealised value” – a gap the company is trying to close with Claude.
Jo Zhu Kennedy, who heads startup platforms at the American AI company, outlined three pillars through which Claude seeks to deliver value to enterprises: building smarter employees, accelerating processes and enabling transformative products.
Carly Ryan, a member of the technical staff in Applied AI in Anthropic, noted that startups are already integrating AI in processes in functions such as marketing, product management and legal.
For many enterprises, AI adoption comes down to cost and flexibility, but the prospect of being locked into a single model has been flagged as a concern.
“When you talk to startups in 2026, what matters most to them is not betting on a single model,” said Kennedy, noting that Anthropic has launched multiple models since 2023.
It is why Claude takes a multi-model approach, which routes tasks dynamically based on complexity, she added. Simpler prompts are handled by lighter, more cost-efficient models; more demanding tasks are assigned to larger models serving an advisory function.
However, as capabilities advance, regulatory scrutiny is also intensifying.
Most recently, the company’s latest AI model, Mythos, raised concerns among regulators over its ability to identify cybersecurity vulnerabilities and its potential misuse.
In response, the Monetary Authority of Singapore urged banks to strengthen their safeguards and plug gaps in their cyber defences.
