Jan 27, 2026
Making Every Employee a Skilled User of AI: Sony Group’s Efforts Toward “AI Democratization” in Enterprise
On December 2, John Kodera, Sony Group's Chief Digital Officer (CDO), took to the stage as a speaker at AWS re:Invent, Amazon Web Services' (AWS) annual conference on AI and cloud computing, held in Las Vegas, USA. Kodera was a speaker accompanying the keynote speech given by Matt Garman, CEO of AWS.
The Sony Group is pursuing a number of initiatives based on the idea that AI supports people. Since 2023, the Group has been an early adopter of AI in the enterprise domain, with the aim of improving corporate productivity and democratizing AI. In July 2025, the Sony Group launched the AI Acceleration Division to lend further impetus to these efforts. In this article, we interviewed Masahiro Oba, head of the AI Acceleration Division, to discuss the background behind the division's establishment and the progress of AI utilization in the enterprise domain.
-
Masahiro Oba
In 2002, Masahiro Oba joined Sony Corporation (now Sony Group Corporation), as a systems planner in charge of information management systems, and B2E reform and also worked as CIO staff. After being posted to Sony Europe (UK) as an IT architect, he led workstyle transformation and Group structure reform initiatives.
He was appointed general manager of the CIO Office in 2019, and general manager of Group Governance in 2021, has been leading the deployment and adoption of AI for the Sony Group since 2023.
The AI Acceleration Division was established in July 2025 to further promote the Sony Group’s AI transformation and democratization of AI.
Getting the Most Value out of AI for Enterprise Use Case
──First of all, please tell us your view on AI.
I do not regard AI as anything "special." This is because I believe that AI is just one of many other technologies that support people. Since AI is a technology based on the accumulation of human-generated data, it does not create value from scratch by itself. Rather, it supports humans and draws out their capabilities.
However, one new value that AI or large language models (LLMs) offers is that it enables users to interact using natural language. And like personal computers and cloud technology, I believe that this value is created only when humans fully utilize it on a large scale.
──Is that the thinking behind the "democratization of AI in enterprise"?
Yes, that's right. In discussions with management about the company’s fundamental stance on AI, we agreed that it is critically important for all employees to understand both the opportunities and risks of AI in the enterprise setting. As a company that respects technology, we want all Group employees - including those in non-engineering positions - to be exposed to AI, to know not only its potential but also its risks, and to be able to use AI. This is our goal, and what is meant by the "democratization of AI in enterprise."
Ensuring Safe and Secure Usage of AI
──Can you update us on how the democratization of AI in enterprise is currently progressing?
The turning point was the explosion of ChatGPT in 2023, which made AI easily accessible to the general public. We then reevaluated how the Sony Group should approach AI for enterprise, and came to a consensus on the aforementioned direction that "the most important thing is for all employees to recognize the opportunities and risks of AI," and began working toward the democratization of AI in enterprise.
At this point, two and a half years later, approximately 58,000 active users in the Group are performing 150,000 inferences every day in the enterprise domain. In addition to chat-style use, an increasing number of AI applications for enterprise are also active in background processing through APIs. With 380 business Proofs of Concept in the preliminary stages of incorporating AI into day-to-day operations and more than 70 cases of AI used in enterprise operations, AI is becoming more common in a variety of areas.
The AI Acceleration Division was established in July 2025 to further advance the democratization of AI in enterprise, including exploring the future vision for operations at Sony Group Corporation, which serves as the Group’s headquarters.
──What are the challenges in democratizing AI in enterprise?
A particularly significant challenge is the management of the various risks. From the outset, concerns were raised about information leakage and potential infringement of intellectual property rights. In addition, hallucination - the generation of incorrect information - was identified as a critical issue. Recognizing that business adoption would be impossible if these issues were left unaddressed, our priority has been to rapidly establish a system that enables all employees to use large language models (LLMs) with confidence.
──That was the Enterprise LLM that you released in August 2023. Can you explain Enterprise LLM in more detail?
Enterprise LLM is a chat-type application that enables Group employees to freely select and use their preferred LLM, supported by contracts with various providers offering over 140 of the latest LLMs. The platform on which Enterprise LLM is based can operate without downtime, even when used by a large number of employees, and is cost-effective because of the Group-wide umbrella agreement that leverages the scale of the Sony Group. In addition, since the input data is securely contained within the Sony Group, it is widely accepted within the Group as a safe and reliable LLM solution.
At the beginning of the project, the team brainstormed many potential names, but ultimately chose "Enterprise LLM." The name clearly communicates its purpose - an LLM designed for companies and enterprises. While simple, it conveys a strong and reassuring message of being a safe and secure solution that can be confidently used within enterprise environments.
──How did you achieve safety?
Taking the need to prevent information leaks as an example, we not only ensured platform encryption, but also implemented safeguards with LLM providers to ensure that any data entered into the AI remains secured within the company. When introducing AI, we also scrutinized the underlying technology and data as well as internal rules and use cases, such as scenarios in which AI should not be used.
We also had specialists in security and legal affairs join the team from the outset to ensure that legal and ethical concerns were cleared, and to create an environment in which AI could confidently be integrated into business operations.
Of course, the appropriate use of AI within a company requires internal assessment and ongoing review, as well as measures to address the economics and information security. To this end, the Sony Group has established a rigorous approval process involving departments such as Legal, Data Privacy, Security, and AI Governance (a mechanism to ensure ethics, transparency, and fairness), which enables the "safe and secure" introduction of the latest models to Enterprise LLM.
Accelerating Awareness
──In addition to creating a safe and secure environment, what is important in achieving the democratization of AI in enterprise?
In parallel with creating the environment, we prioritized awareness initiatives. Dedicated teams were established to design and deliver training sessions and events across the Group, both in Japan and abroad. Beyond introducing the latest features, we promote the adoption of AI usage at all locations by sharing practical use cases from various departments.
──What are you mindful of in your dissemination of information?
We are mindful of the extraordinary speed of AI’s evolution. The biggest concern is that employees who had been exposed to AI six months earlier would conclude that AI was not yet ready for business use, and that this perception would persist. This is precisely the kind of situation that must be avoided in this ever-evolving field, where the situation could change completely in the span of a month.
We therefore ensure that the latest updates are delivered promptly. We regularly share timely information, such as "Here is what you can do now," or "This is what’s new in the latest models." In addition, we create multiple engagement opportunities, including e-learning programs and in-person events.
When sharing case studies, we make it a point to include failures as well. In the field of AI, many things that could not be done before can be done six months later. We believe examples of failure are equally valuable, and that actively communicating how a past failure evolves into success conveys a sense of speed and progress.
Sony's Use of AI in Enterprise: Where We Are Now
──What kinds of use cases are emerging as the democratization of AI in enterprise progresses?
Departments have begun to use the system to communicate with users, and a wide range of examples have emerged, including information retrieval and analysis, knowledge transfer, business document analysis and review, and security verification. Since the Sony Group conducts business globally, there are many examples of applications that go beyond mere translation, such as translations that reflect regional and business characteristics.
──What do you see as the reason why the use of AI in enterprise has progressed so well within the Sony Group?
One key factor, I believe, is the close discussions we had with senior management, which allowed us to advance the project with shared awareness of the issues and aligned expectations. Additionally, the clear messages communicated at key moments by management served as a major driving force.
We also put careful thought into team setup. Top talent from various divisions within Sony Group Corporation have been invited to participate in the project in concurrent roles. It was also very significant that we were able to create a system where future leaders can autonomously advance the project and dedicate themselves to their work.
People Continue to Play the Leading Role
──Will advances in AI bring changes to people and organizations?
If we go back all the way to the start of the Industrial Revolution, new technologies have often changed the expectations placed on humans. I feel this applies to AI as well. I believe that there are areas where AI will increase efficiency, and society as a whole will likely invest in these areas. This may lead to the question of whether AI will replace humans, but I don't think that is the case.
As I mentioned in the beginning, AI is merely a technology built on the accumulation of human-generated data. It is always people who create new value, and the Sony Group places great importance on this concept. We believe that because we are in an environment where many employees are using AI in enterprise, we can create a "people-oriented" organization that strategically utilizes AI. This idea is also important in our ongoing considerations for the ideal vision for the future operations of Sony Group Corporation.
──What is your vision for the future of the Group after the democratization of AI in enterprise?
When I was asked to head a team that would promote the democratization of AI in enterprise in 2023, I thought, "I don't want this department to still exist five years from now.” Once the use of AI in enterprise becomes the norm, as with today's PCs and cloud computing, this team will end its role of promoting the adoption of AI within the company, and move on to the next challenge. That is the goal I aim to achieve.
Sony’s Purpose is to “Fill the world with emotion, through the power of creativity and technology.” Five years from now, I don’t want us to be a company that prides itself on ‘advanced AI utilization.’ Instead, I want us to be a company where human creativity flourishes and continues to deliver new and inspiring experiences and emotions. And behind the scenes, AI in enterprise is being fully leveraged - that is the ideal state.
