Japan – notoriously slow to adopt digital technologies common in the developed world – has become a surprisingly quick adopter of AI as it faces both a shrinking population and an aging digital infrastructure based on outdated code.
“Japan was our top or second most popular country in terms of overall user engagement,” Russell Kaplan, president of Cognition AI, the San Francisco startup behind AI coding tool Devin, said in early June.
The East Asian country has the oldest population in the world, with almost 30% of residents over 65 years old. Japan’s working-age population is expected to decline increase by over 30% by 2060. The decline leads to a shortage of programming talent: in 2023, Japan’s Ministry of Economy, Trade and Industry (METI) estimates that the country will There is a shortage of 789,000 software engineers until 2030.
Cognitive AI is Make Japan the first step in expanding into Asia: opening an office in Tokyo in April; Later this year, Singapore will become its Asia Pacific headquarters.
The company believes Japan will be the ideal testing ground for AI-powered software development. “The need is real, particularly in critical infrastructure and government,” Kaplan said. “The country has aging infrastructure and a declining workforce.”
The efficiency gains could be immense. Faced with a national IT compliance requirement, the Sapporo city government needed to modernize over a million lines of legacy code, which Kaplan estimated typically took 200 months of development work. With help from Devin, Sapporo engineers completed it in about a quarter of that time.
Even before Cognition was officially launched in Japan, Devin had already gone viral in Japan. “There was a debate about what the proper honorific was for Devin,” Kaplan said, referring to the name suffixes used to denote social hierarchy.
“The community agreed on this Devin-kun.”
Japan’s bet on US AI
Japan has become the preferred beachhead for U.S. AI companies seeking global expansion. OpenAI and Anthropic both opened their first international offices in Tokyo. Microsoft, alphabetand other hyperscalers have Billions pledged to Japanese data centers. Japan also became the second country to secure access to Anthropic’s powerful Mythos model with three of its largest banks – MUFG, Mizuho and Sumitomo Mitsui – among others who were granted access through Project Glasswing, a program designed to help key companies and critical institutions address security vulnerabilities. (This access was switch off quickly after the USA banned all foreigners from using the model in mid-June.)
While South Korea, Singapore and other regional economies have prioritized sovereign AI, Japan appears more comfortable sticking with U.S. AI due to the country’s investments and close relationships with American AI labs.
Courtesy of Cognition AI
“Japan has invested disproportionately in working closely with U.S. companies to influence those companies’ roadmaps and meet local domestic needs,” Kaplan said. One of OpenAI’s largest investors is Softbank, the giant Japanese telecommunications company that OpenAI runs Tech booster Masayoshi Son.
AI could represent an opportunity for Japan to integrate its digital systems with the rest of the world. Kaplan suggested that low English proficiency “has led to a little more isolation at some companies in Japan.” But the native multilingualism of the AI overcomes this barrier. Japanese engineers can use Devin to work entirely in Japanese while collaborating with teams on the other side of the world through the agent.
AI coding reaches Asia
Founded in 2023, Cognition AI is best known for its AI coding tool Devin. The tool acts as a complete AI software development teammate: give it a task, and it autonomously programs, debugs, and deploys code into the tools a development team already uses.
Devin was one of the first proponents of “AI workers,” or agents, which are fully integrated into workplace tools like Slack and to which employees can assign tasks without having to resort to constant prompts.
At the end of May, Cognition increased more than $1 billion in a new round of financing This valued the startup at $26 billion, more than double its valuation from the September 2025 round. The company’s annual run rate reached $492 million at the time of the raise, up from just $37 million a year earlier.
Cognition AI’s coding tools pose, for some investors, an existential threat to existing programmers and software developers, particularly in countries like India, a traditional hub for back-office work. The prospect that AI agents could do the same work at a fraction of the cost has unsettled investors. Shares in Infosys, Wipro, Tata Consulting Servicesand HCLTech are all down 30% to 40% in the last 12 months.
But Kaplan isn’t worried about India’s ability to adapt to AI. “On the ground in India, an engineer’s job can be more fun and impactful. Suddenly you have someone who has been working alone on a particular part of a project and they get a promotion where they have a whole team of AI agents working for them.” Kaplan said. “The companies we work with are using productivity improvements to become more ambitious.”
One of Cognition’s more unexpected growth markets is Malaysia. The country’s capital, Kuala Lumpur, has emerged as a regional software development hub due to a large English-speaking talent pool, lower operating costs and proximity to the rest of Southeast Asia. Kaplan described the engineers his team encountered there as the most experienced in the world at managing AI agents.
Cognition has launched an Applied AI Engineering program in Malaysia that identifies top engineers who excel at leading agents and trains them to teach entire teams how to work effectively with AI.
Kaplan is also closely examining South Korea and Australia as possible expansion markets in the Asia-Pacific region.
Cognition’s growing presence brings to light another advantage. Computing power, the computing power at which AI systems run, is a finite resource; According to Kaplan, demand at Cognition doubles about every seven weeks. But geographically dispersed teams allow computing power to be used during off-peak hours on Wall Street and Silicon Valley. “When people are working in Japan, people are sleeping in New York,” Kaplan said. “When you work this way as an AI company, you get a lot of efficiency.”