U.S. companies are taking an increasingly international view of their AI ambitions, with OpenAI this week opening an office in Tokyo and Microsoft firmly planting its flag in the Middle East with a $1.5 billion investment in G42, an AI company in the United Arab Emirates (UAE).
Both moves are aimed at getting footholds in a rapidly expanding global AI market – which researchers with Statista say will expand from about $184 billion this year to $826.7 billion by 2030 – with Microsoft’s deal also helping the United States’ geopolitical ambitions of hindering China’s expansion in the Middle East.
U.S. Commerce Secretary Gina Raimondo reportedly had a hand in the Microsoft-G42 agreement.
For Microsoft, the investment in G42 is an expansion of a year-long relationship with the UAE company and an important steppingstone to other regions. G42 will run its AI applications and services on Microsoft’s Azure infrastructure and partner with the IT giant to deliver AI solutions to public sector organizations and large enterprises throughout the Middle East, Central Asia and Africa, according to Microsoft.
The deal “will enhance the UAE’s position as a leading AI hub and provide digital infrastructure and services to help underserved nations innovate and grow,” Microsoft President Brad Smith wrote in a post on X (formerly Twitter). “We will combine world-class technology with world-leading standards for safe, trusted and responsible AI, in close coordination with the governments of both the UAE and the United States.”
Peng Xiao, group CEO of G42, said in a statement that Microsoft’s investment, will help the company’s efforts to “deliver cutting-edge AI technologies at scale. This partnership significantly enhances our international market presence, combining G42’s unique AI capabilities with Microsoft’s robust global infrastructure.”
Keeping China Out of UAE
According to reports, G42 will remove equipment from China-based Huawei – a company that U.S. policy and lawmakers said is a national security risk due to its close relationship with the Chinese government – from its infrastructure. This follows G42 divesting from other Chinese companies, including ByteDance, the parent firm of controversial video-sharing site TikTok.
Adding more to the geopolitical machinations behind the Microsoft-G42 partnership, the companies assured the U.S. and UAE governments in a “first-of-its-kind agreement” they will use best practices to ensure the secure and responsible development and deployment of AI, a key concern of the Biden Administration and the focus of the President’s executive order released in October 2023.
Microsoft and G42 agreed to enhance the security and compliance framework of their shared infrastructure and comply with both U.S. and international trade, security, business integrity and responsible AI regulations. The companies, in consultation with their governments, outlined their intentions in an Intergovernmental Assurance Agreement, according to Microsoft.
The two also will create a $1 billion development fund to expand the AI talent pool in the Middle East.
OpenAI Lands in Japan
As part of OpenAI’s expansion into Asia, the company named Tadao Nagasaki as president of OpenAI Japan and are giving local businesses early access to a GPT-4 model optimized for the Japanese language.
“This custom model offers improved performance in translating and summarizing Japanese text, is cost effective and operates up to 3x faster than its predecessor,” OpenAI Sam Altman wrote in a blog post. “Speak, a top English learning app in Japan, is seeing 2.8x faster tutor explanations in Japanese when users make a mistake with a 47% reduction in token cost, unlocking higher quality tutor feedback in more places and with higher limits per user.”
OpenAI is releasing the custom model more broadly in the API later this year.
The company has other international offices in Europe, but Tokyo is its first in Asia. Altman wrote that a presence in Japan also locates the company to such companies as multinational Daikin, technology conglomerate Rakuten and Toyota Connected, a company launched in 2016 to expand Toyota’s mobile tech capabilities. It uses OpenAI’s ChatGPT Enterprise tool for such jobs as automating business processes, help with data analysis and improving internal reporting.
Yokosuka City also uses ChatGPT, giving most city employees access, he wrote.
At the same time, Altman and other OpenAI executives also are looking to strike up partnerships with Japanese semiconductor companies and find new sources for the GPUs that are crucial for training generative AI models. Nvidia is the GPU kingpin, but demand for such key GPUs like its H100s has been high, and availability hasn’t always kept up.
In an interview with Japanese news outlet Nikkei, OpenAI COO Brad Lightcap said it’s important to expand the number of sources and avoid shortages because demand is only going to grow,
“Our priority is making sure that we don’t end up in a world where there’s so much demand for AI, but we don’t have enough capacity,” he said.
OpenAI has considered building its own chips to offset the tight market, and Altman – who for several years has been talking about the need for more global compute capacity – at one time was reportedly talking with Japanese company SoftBank and G42 and the UAE government about raising money for fabs, telling them he needed to raise trillions of dollars for such an effort.
