"Artificial intelligence (AI) is not just a tool. It is the engine that drives the mobility ecosystem."
Mike O'Gle, head of public policy for Uber Asia-Pacific (APAC), presented on "Uber and AI, the restructuring of global mobility" at the SMARTCLOUD SHOW 2025 held at the Westin Josun Hotel in Seoul on the 27th. He noted, "AI is coordinating global movement by executing billions of predictions every day," adding, "Regulations should be created based on risks and harm, not on the technology itself."
O'Gle introduced Uber's global achievements, stating that the company handles an average of 33 million rides daily and has provided more than 61 billion movement services to date. The monthly active users (MAU) number 180 million, with 8.8 million drivers and delivery personnel, and the number of partner stores is 1 million. He explained, "The income returned to the community through this ecosystem has accumulated to $350 billion."
He emphasized that AI is the key to managing such a large scale in real-time operations. O'Gle stated, "Uber trains or retrains 20,000 AI models every month," revealing that these models perform 10 million predictions per second and 20 billion predictions daily. He added, "These predictions are not just simple simulations; they reflect every step of platform operations, including demand-supply matching, optimal route calculation, pricing, and risk detection."
He also highlighted examples of how AI enhances safety and user experience in the food delivery service "Uber Eats." O'Gle said, "When a delivery partner takes a selfie while wearing a helmet, AI automatically verifies it to ensure safety," adding that features consumers use daily, like food recommendations and automatic destination marking, are actually backed by billions of predictions.
To this end, Uber has established principles for responsible AI use. O'Gle mentioned that "Uber has set six principles: establishing governance systems, securing reliability, verifying fairness, ensuring transparency and explainability, maintaining quality, safety, and security, and protecting personal information," explaining, "We have a dedicated organization in-house to check for model biases and unfairness."
O'Gle pointed out that as regulations in various countries are increasing, the corporate environment is becoming more complex. He noted, "In 2015, there were no AI regulations in the U.S., but now there are over 450," and added, "The patchwork situation that varies from state to state and city to city is a significant burden for global corporations." This year, the U.S. announced its "AI Action Plan," focusing on job creation, minimizing biases, and preventing misuse. In contrast, the European Union (EU) adopts a risk-based pyramid approach, strictly regulating high-risk areas like biometrics, critical infrastructure, health, and finance while managing low-risk areas like food and movie recommendations more simply.
O'Gle further emphasized, "Singapore is leading model governance, India is establishing industry-specific standards, and Korea and Japan are focusing on enhancing transparency and safety," adding, "Regulations should be designed based on risks and harms rather than restricting specific technologies." He continued, "We need to prioritize management in clear high-risk areas such as child safety, health, and finance, and respond according to the context of countries and industries."
O'Gle stated, "AI is not simply a tool for creating images but an operating system and engine that coordinates urban movement and commerce in real-time," stressing, "As the engine gets stronger, safety and accountability regulations must also be strengthened. We will actively participate in global regulatory discussions while establishing our own governance."