Television, please. Institute for research. Societal impact of AI on the world, into future risks we may face, such as cyber, bio, and into the economic impacts of this technology.
And my job and the work that we do is that we are sharing information with the world about what AI is doing today and what it will do in the future. I am a co-founder of Anthropic and head of the benefit department. I also lead the Anthropic Institute. We primarily conduct research here, investigating the global impact of AI on society, as well as future risks including cyberattacks, biological threats, and the economic impacts this technology may bring. My job is to share as much information as possible with everyone, focusing on what AI is doing today and how it will evolve in the future.
Among those roles, I consider the most important to be communicating how AI and coding will expand and be used in fields like AI for science and AI for robotics. In AI systems to get better. Transform the economy a lot more broadly, and Japan is especially well-positioned here due to expertise in the AI supply chain, in the creation of scientific equipment and robotics. Japan is set up, supercharged by this big technology trend.
Looking ahead to the next year, I would like to touch upon how coding will change AI systems and discuss two specific areas where further development of AI systems is expected. AI is expected to transform the economy in various ways, and I believe Japan will play a very significant role in this process. For example, Japan possesses broad expertise in areas such as AI supply chains, scientific equipment, and robotics. In that sense, we believe this technology will supercharge Japan moving forward.
Furthermore, as Anthropic, we want to expand beneficial collaborations within Japan. We have already begun initiatives with the University of Tokyo in the form of a Japan AI impact index. This is an initiative to create a Japanese version of our Economic Index, further tailored to the current situation in Japan by reflecting Japanese data. Through this, we believe that when the government makes decisions in the future, for example, they will be able to do so with an understanding of the impact AI has on the economy and society.
Our team members in Japan are present, and we look forward to expanding our very deep partnerships with central companies here, and our team also includes members like Mr. Akita and Mr. Clifford who can assist with policy in the field of cyber defense. As Anthropic, we believe we are reaching a critical juncture over the next few years, and issues such as those related to AI and security are topics that will be indispensable as we continue our development. We believe this will also be an important challenge in Japan, and with that, I would like to conclude my remarks and open the floor for any questions you may have later, so please feel free to ask.
Anthropic Japan Representative
6:05 ↗
Hello, everyone. Thank you all for gathering here today. I would like to briefly introduce our current status and look back on the past year. Yes, this one. This is it. As we celebrate the first anniversary of Anthropic Japan, a year spent walking with Japan and looking toward the future, we announced the opening of our Tokyo office last October. A year has passed since then, so I would like to share with you what kind of achievements and events have occurred over this past year.
First, looking back on the year with specific figures. Regarding our team, specifically our headcount at the Japanese subsidiary; back in October of last year, I mentioned at an event that we only had about three employees at the time. Since then, we have grown to over 50 members, and I feel we are now able to engage with customers and partners across a much broader scope.
Second point. As for the number of enterprise adoptions, while I cannot disclose specific figures, as a benchmark, we now have more than 10 times the number of customers utilizing Claude compared to this time last year. Finally, regarding our usage index, this is cited from an economic index. This index measures how many people are using it relative to the population. If the global average is 1.0, Japan is at approximately 1.92, which suggests that the adoption of the AI Claude in Japan is quite advanced.
Yes. Moving to a more qualitative review, as the foundation of our Tokyo office activities, expanding our staff is crucial for us to communicate directly with you. We have focused on hiring for sales and pre-sales, which we refer to as Applied AI at our company. We have also made progress in hiring architects, product engineers, partnership managers, as well as roles in marketing and corporate functions, which has enabled us to engage in meaningful conversations with our Japanese customers. I believe this is a significant achievement.
And the second point is that I believe adoption among enterprises has progressed quite significantly. Regarding our partnerships, as many of you have seen in the press, we were able to establish strategic partnerships with major system integrators earlier this year, including NEC, Hitachi, and Fujitsu, as listed here. Furthermore, these system integrators are in a position where they don't just use our technology internally, but work alongside us to propose AI solutions to customers across Japan.
At the same time, we are seeing significant progress in adoption among those customers, the user companies themselves. I will introduce this in a later slide, but the industry reach is becoming very broad as well. And for the third point. This is a very important initiative because our mission is to provide safe and secure AI; we are actively engaging in various efforts to become a trusted partner in Japan for AI safety and policy development. As for what I can share, we signed an MOU around this time last year with the Japan AI Safety Institute, and building on that foundation, we are currently providing our cooperation in the evaluation of AI. Also, as Jack mentioned, we are collaborating with a research laboratory at the University of Tokyo on the Japan AI Index.
This slide serves as a summary of the previous two, so I will skip the overlapping parts, but to mention something I haven't introduced yet, we held a "Code with Claude" event ourselves in Tokyo back in June. It was primarily a venue for developers to gather, where we introduced our cutting-edge technology and shared use cases, and we had some announcements to make there as well.
For one thing, we truly value our community, so we took that opportunity to call for people to become our ambassadors. In just two days, we received thousands of applications, and we now have several dozen individuals selected from that group who are actively involved in our community activities. Furthermore, as has already been announced, we are moving forward with various initiatives, such as our AI collaboration with the SBI Group and our work with NEC and other financial institutions.
And looking at the most recent three to four months, we launched Project Glasswing, which involves providing access to our Mythos model. Yes. During this period, we have also begun providing some Japanese companies with access to that model. Around that same time, we released a new model called Phaigle, which sits above our Opus model, which was our most advanced model at the time.
Paul Smith, our Chief Commercial Officer, who is in charge of sales and marketing, also visited Japan, and I believe it was a period where we were able to have very productive conversations with our Japanese clients. Also, as you may know, we announced a partnership with the Mori Art Museum, and this October marks the first anniversary of that partnership, with an exhibition by Mariko Mori set to begin. We are serving as a sponsor, and Claude is being used in the production of various works, so I believe the utilization of Claude is becoming quite widespread.
Also, Hitachi recently announced the CIDP, or Critical Infrastructure Defense Program. Details are available on the website, so please take a look. In Japan, Hitachi is participating in this initiative to strengthen the defense of vital social infrastructure, particularly against cyberattacks. And we have Jack from Rutter here with us today, don't we? Yes.
Furthermore, regarding our Japanese customers and how they are using it, we haven't often had the chance to present these logos to everyone, but today we are showing them in this format for the first time. The number of companies has increased more than tenfold compared to last year, so we cannot list them all, but we have included a selection of customers who have permitted us to use their logos here. I won't call out every single name, but since you are taking photos, I hope you will take a closer look later.
I would like to introduce just two; at the bottom is Digital Hollywood, and it is being announced that they are the first university in Japan to provide access to Claude for all students and all faculty members. That is the first one. And another one is the Digital Agency; although their logo isn't shown here, we have received permission to speak about this today. As many of you likely know, the Digital Agency has a Government AI initiative, and we are pleased that their staff can now access and use Claude as the AI model within that framework. Yes. I believe those represent some of our major clients and achievements.
Meanwhile, regarding our partner companies, we are significantly expanding those collaborative relationships, including with the partners I mentioned earlier. We are looking to expand this even further. In Japan, in particular, companies often rely on the expertise of system integrators for building systems and introducing AI. Given this history, our strategy is to work closely with these partners to deliver AI—to deliver Claude—to companies across Japan. Yes.
Furthermore, as some examples have already been shown, we are involved in various initiatives beyond the corporate sector as well. We have already introduced the AI Safety Institute, and in addition, a new exhibition will begin at the Mori Art Museum starting on October 31st. At the University of Tokyo, we have the Japan AI Index, and in another area we haven't introduced yet, we have become a sponsor of the basketball team, the Tokyo Z. We are supporting the players using Claude, and those efforts are already underway. Yes.
This is a summary of what we have been working on over the past few months. In the field of cybersecurity, we have begun providing Japanese companies with access to MITRE ATT&CK. This isn't everything, of course, and since there are various ways to utilize MITRE, we hope that Japanese companies will make good use of it in their cybersecurity contexts, covering a broad scope.
And one more thing, this is something we recently announced—it's a mechanism to support physical AI—we have begun offering a preview version of the Model Hardware Standard, or MHS. This acts as a framework for connecting hardware control, in particular, with AI. We believe this will be a major step forward for the future of physical AI. Once we begin to see concrete results in this area, we would like to share them with you all as well. At this moment, it is still in preview. Yes.
And finally, for these last two slides, as we enter our second year, we would like to focus on these five points under the themes of "expanding" and "deepening." Well, of course, the first point is the full-scale adoption in corporate and public sectors, and the second is the strengthening of our partner business. As mentioned earlier, usage among our partners is progressing, and we want to work firmly alongside them to deliver AI to Japanese companies.
Third is the use cases. There are many use cases, but we want to uncover those that are unique to Japan or unprecedented globally, especially in sectors like manufacturing where Japan is strong, to see how we can find new ways to apply AI in those environments. Fourth, as I introduced in several slides, we want to maintain strong touchpoints with research, policy, and culture, and apply these to the social implementation of our safe AI.
Finally, we plan to expand our hiring to grow our Japan team and further strengthen these activities. Yes. This is our final slide. We are committed to building trustworthy AI together with everyone in Japan. Thank you for listening. Thank you.
