Since around July 2026, I have basically stopped using Word to write papers. The main reason is that typing footnotes in by hand, one by one, was a real chore. Since I take notes while reading the literature anyway, what I wanted was for the tool to place the right footnotes from those notes automatically and tidy up the citation style as well. On top of that, when I ask Claude Code to read a manuscript and check it for typos and the like, having it written as a Markdown file is simply more convenient. And, quite simply, Word is sluggish and its UI is not very pretty, so as a tool to spend long hours with it had always bothered me (apps built with the Tauri framework are light and look clean).

In short, this is how writing with Genkōdai works. I collect excerpts from what I read in notes organized by topic rather than by paper, and each excerpt is written as a set: the source, the quotation (verbatim), and my own comment. When I need a source while writing the main text, I don’t go looking for it on the spot; I just drop a “citation needed” marker and keep going. Later, when I click the marker, the excerpts for that topic are listed as source-and-quotation pairs, so I can look at the original wording, pick the right one, and replace the marker with a footnote at the press of a button. If there is no matching excerpt, it simply shows “none found” and never makes up a plausible-looking candidate.

I decided to release it because, to someone who doesn’t know what actually goes on, writing papers in Markdown could look exactly like having Claude Code generate the whole paper. All the more so because, if you look at the properties of a Word file exported from Genkōdai, the editing time shows up as something like one second… So I thought that properly publishing both the idea and the technical specification might also be a way of protecting myself.

In fact, what I was most careful about in building Genkōdai was not letting AI create sources. There is a feature that lets Claude Code search for footnote candidates in one go, but Claude is only given permission to read files and cannot rewrite the manuscript. It may search only within the excerpts I have taken myself; whenever it proposes a candidate, it must copy the excerpt’s original wording exactly, and the app mechanically checks whether that wording really exists, character for character, in my notes, discarding anything that doesn’t match. “None found” is treated not as a failure but as a finding that there is still literature left to read. I decide one by one whether to accept each candidate, and a history (git) is recorded automatically before and after each run, so every change can be traced afterwards. It is closer to the truth to call it a tool with a mechanism built in to keep AI from writing than a tool for getting AI to write papers.

The screenshots in this post use a sample manuscript made for publication.

That said, since I built it around my own way of writing, it requires that way of working as is, including how the notes folder is organized, and for now it only runs on Mac and Linux. There is no installer either, so to use it you need to build it from source. I would like to do something more about the design and the name, and I hope to improve it at some point, including Windows support. The license lets you use and modify it freely for non-commercial purposes (PolyForm Noncommercial). The design documents are published alongside it, so you are very welcome to borrow just the approach rather than the tool itself.

Building an editor that fits my own way of writing, makes the work more efficient, and properly harnesses AI so that it isn’t pushed into unwarranted inference was great fun and a worthwhile experience.

Presentation · 2026 Korea-Japan Competition Law Academic Research Conference (Korea University)

On September 18, I gave a presentation titled “Recent Developments in Japanese Competition Law: Through the Lens of Innovation” at the “2026 Korea-Japan Competition Law Academic Research Conference (2026 한일 경쟁법연구 학술교류대회),” held at Korea University in Seoul. The conference was co-organized by the Korea Competition Law Association (KCLA) and the Japan Association of Economic Law, and doubled as an exchange between Korean and Japanese economic law scholars.

I was originally asked to speak on “recent developments in Japanese competition law and policy.” But an exhaustive survey of Japanese developments would not make good material for discussion among researchers, so I chose to examine Japanese competition law enforcement critically through the lens of innovation. I chose innovation because it covers recent new issues while also giving a cross-cutting view of the whole of enforcement.

The presentation builds on my article in Kosei Torihiki no. 892, “Strategic Conduct Aimed at Excluding Competitors and Innovation”, and my research on predatory innovation presented at the Japan Competition Law Forum.

In antimonopoly law, innovation appears both as a factor in assessing the legality of individual conduct and as a goal (or slogan) of competition policy. I organized the presentation along these two axes.

For the former, I reorganized the discussion not by type of conduct but by “which question we are trying to answer”: how to distinguish legitimate innovation from unlawful exclusion; with what probability to anticipate uncertain long-term harm; and how to weigh short-term harm against long-term benefits. Divided this way, it becomes clear that Japanese cases and debate are heavily skewed across the questions. The first question has a small number of important cases, such as the printer litigation over technological tying; the second has no cases; and for the third, a framework has been set out only in the context of merger control. On the second question, I also took up U.S. debates that have not yet been sufficiently introduced in Japan, such as Hemphill and Wu on acquisitions of nascent competitors and Salop on designing presumptions and burdens of rebuttal from error costs.

For the latter, I took up the JFTC’s statement of January 2026, “Proactive Development of Competition Policy for the Promotion of Innovation,” and the guidelines on transactions in intellectual property of June of the same year. I pointed to a structure in which, under the framework of abuse of a superior bargaining position, “hindering innovation” is asserted as a harm, yet proof of harm at the market level is never required.

Finally, I took up remedy design, comparing the JFTC’s Google order and MC Data Plus order with the final judgment in the U.S. Google case. Japanese law incorporates effects on innovation into its theory of harm when assessing legality, yet at the remedy stage it is not designed to adapt to technological change. I presented this asymmetry as a challenge for Japanese law.

It was a very stimulating conference and, I think, a wonderful meeting that promises more active academic exchange between Korea and Japan. My sincere thanks to the Korean and Japanese colleagues who worked to plan and run it, and to everyone who asked questions and offered comments.

Presentation · Monash Warwick Alliance Digital Regulation Workshop (Monash University)

I was too busy to publicize it in time, but on August 14 I gave a presentation titled “Rules, Standards and Observability under Japan’s Mobile Software Competition Act” at the Monash Warwick Alliance Digital Regulation Workshop, held at Monash University in Australia. The other speakers were all there in person; I was the only one joining over Zoom.

The workshop considered, from a competition law perspective, the future of ex ante regulation of digital markets in Australia in light of the experience of the EU, the UK, and elsewhere. Australia is in the middle of designing its own ex ante framework, and I was asked to talk about how Japan’s experience might be drawn on.

My material was the roughly seven months of operation of the Mobile Software Competition Act (the Act on Promotion of Competition for Specified Smartphone Software), which came fully into force in December 2025.

Looking at the domestic debate since it came into force, the controversy is concentrated on the level and basis of Apple’s and Google’s fees. The argument over whether fees on alternative payments and external links effectively close off routes to competition takes almost the same shape as the tasks that have been considered hardest in ex post antimonopoly enforcement — justifying costs and assessing effects that exclude competitors. Having introduced ex ante regulation, the biggest dispute is being fought on the same ground as ex post enforcement.

So in my presentation I sorted the Act’s obligations not by “ex ante or ex post” but along two axes — whether an obligation is fully written as a rule, and whether compliance can be observed from outside — into three types. The first is obligations, such as the prohibition on compelling the use of a particular browser engine (Article 8(3)), that are clearly written as rules and whose compliance can be checked from outside. The second is obligations, such as the prohibition on the improper use of acquired data (Article 5), that are written as rules but whose violations take place entirely inside the firm and leave no outward trace. The third is obligations, such as those involving fees, where room for assessment as a standard effectively remains within the ex ante regime.

Seen this way, my argument is that whether ex ante regulation works depends not on the dichotomy of ex ante versus ex post but on whether each obligation could be fully written as a rule, to whom its compliance is observable, and whether areas where standards remain are equipped with information mechanisms to support them (such as the granularity of compliance reports, and the allocation of disclosure of and accountability for the basis of fees). Rather than taking fee issues out of ex ante regulation, one should add devices that bring the information needed for analysis to the regulator’s side — a point I think may be useful for Australia as it builds its own system.

Debates on ex ante regulation tend to become debates about the strength of regulation: hard DMA-style rules or transparency-based co-regulation. What distinguishes this presentation is its focus on rather unglamorous design variables: rules versus standards, and observability.

The discussion paper is still a draft, so I will revise it in light of the comments I received at the workshop and hope to turn it into an article eventually. My thanks to the organizers for inviting me.

Book · Yuhikaku (eds. 稲葉僚太, 大泉玄之助, 伊永大輔, 中島菜子; contributed chapters)

The co-authored book Competition Policy in the Platform Era: At the Cutting Edge (Dejitaru purattofōmu o meguru kyōsō seisaku no saizensen), edited by 稲葉僚太, 大泉玄之助, 伊永大輔, and 中島菜子 (Yuhikaku), has been published.

For each issue and case concerning digital platforms, the book sets practitioners’ views side by side with commentaries by legal scholars. There are also chapters of commentary by economists.

I wrote commentaries on the 2024 Google commitment case, the 2025 Google cease-and-desist order (co-authored with Professor Korenaga), the Google/Fitbit and Salesforce/Slack mergers, and four market study reports (digital advertising, mobile OS, news platforms, and connected TV).

Some of the commentaries are written plainly with general readers in mind, but others go fairly deep, for example sorting out the relationship between antimonopoly law and problems that cannot be addressed by mandating opt-outs. I would be really glad if you read it.

Commentary · Kluwer Competition Law Blog · Read the paper ↗

Article · Kosei Torihiki no.904 p.45

Presentation · The IAFOR International Conference on Arts & Humanities in Hawaii (IICAH 2026)

In January I gave a presentation titled “From Data Hunt to Data Vault: A Closed-Chain Solution for Antitrust Compliance” at The IAFOR International Conference on Arts & Humanities in Hawaii (IICAH 2026). The presentation was accepted after peer review, and the conference is interdisciplinary, with many participants from outside law.

The starting point was my article on the Antimonopoly Act and blockchain in NBL no. 1252. That article stopped at raising questions in the abstract, introducing the ideas and background of the technology, so this time I went a step further and built it into a concrete institutional proposal.

The underlying concern is that antimonopoly enforcement is structurally reactive. The relevant data sit with the firms, and the authority has to go looking for evidence only once a violation is suspected (a “data hunt”). The information asymmetry is large, investigations take years, and harm to the market spreads in the meantime.

So I proposed a mechanism (a “data vault”) in which firms in a gatekeeping position record their transactions in advance on a permissioned blockchain — a closed network whose participants are limited to the authority and the firms concerned. Records cannot be altered afterwards, and the authority cannot look at the contents at will: it is designed so that access is possible only when legal conditions are met, such as the opening of a formal investigation. Competitors’ data are kept invisible to one another, and changes to records require the approval of both the authority and the firm, so that neither can change them alone.

My material was the remedies in the U.S. Google (search) case. Remedies such as a ban on exclusive contracts and data-sharing obligations require continuous checking that they are being complied with after the order. I showed that by automatically detecting prohibited clauses when a contract is registered, and by recording whether data requests were answered on time, one can move from periodic audits to continuous verification. As ways of imposing the obligation, I proposed two routes — making it a condition of commitments or cease-and-desist orders, and encouraging participation through incentives such as leniency — and set out a path of starting with gatekeepers and expanding in stages.

Alongside the presentation, I also built a working prototype with Hyperledger Fabric. In a network of four organizations — the authority and three firms — I confirmed that conditional access control and ledger synchronization work. The prototype is introduced as “Complichain” on the Projects page.

Presentation · 2025 Taiwan-Japan Economic Law Conference

I was invited to the 2025 Taiwan-Japan Economic Law Conference and presented on the major developments in Japan’s competition law and policy in 2025, under the title “An Overview of the Major Developments in Japan’s Competition Law and Policy in 2025.”

Presentation · Shojihomu Antimonopoly Law Study Group

At the Shojihomu Antimonopoly Law Study Group, I gave a presentation titled “Designing Remedies in Dynamic Markets: The District Court Decision of September 2, 2025 in United States v. Google LLC.” It was based on the district court’s September 2, 2025 decision on remedies in the U.S. Google (search) case.

I later took up the question of remedy design again in my presentation in Korea, comparing it with Japan’s Google order and MC Data Plus order.

Article · CPRC Discussion Paper no.CPDP-102-J pp.1–31 (in Japanese) · Read the paper ↗

Case note · Hogaku Seminar special issue, Sokuho Hanrei Kaisetsu no.37

Presentation · Japan Competition Law Forum (Unilateral Conduct Study Group)

I was invited to the unilateral conduct study group of the Japan Competition Law Forum and gave a commentary on the cease-and-desist order issued by the Japan Fair Trade Commission in the MC Data Plus case in December 2024. It was the first administrative order the JFTC issued concerning a cloud service.

Commentary · Shin Hanrei Kaisetsu Watch (Economic Law) No.99

Commentary · Kluwer Competition Law Blog · Read the paper ↗

Case note · NBL no.1284

I wrote a case note in NBL on two cease-and-desist orders concerning restrictions imposed by fishery cooperatives and their federation along the Ariake Sea coast on nori producers’ shipments outside the cooperative system.

It is an interesting case in which the main issue was whether a “written pledge” on shipments outside the system, not backed by monitoring or sanctions, constitutes an unjust restriction. In the note I examine in particular the finding of a market foreclosure effect.

The case is now being contested in the district court, and I hope that more detailed findings of fact, going beyond what is stated in the orders, will be made. I would like to write a more detailed analysis once the district court rules, but given the other articles I want to write, whether I can do so is uncertain…

For now, I hope you enjoy it as a brief case note on the orders.

Article · Kosei Torihiki no.892 pp.18–23

My article “Strategic Conduct Aimed at Excluding Competitors and Innovation” appeared in the special feature “Innovation and Competition Policy” in the February issue (no. 892) of Kosei Torihiki.

Since the feature was written by legendary scholars, I felt I had to work hard and write something good. I worked on it desperately for two months, from around October into the new year, and I am attached to it. My second son was born on October 30, so I did most of the research with him in a baby carrier. It was really tough…

I got carried away and drew up a grand research plan, and in the end I had to submit without being able to write everything I wanted. Compared with the ideal I had originally planned, it feels as though I stopped at the introduction of the ideal paper. Someone who could fit all of the intended content perfectly into a six-page manuscript and still make it an easy read would be first-rate…

Even so, this manuscript is the best I could do for now. I would be glad if you took a look.

The “examination from the standpoint of legal hermeneutics” mentioned in the discussion is an issue I spent about a year preparing in 2022 but have not yet published. I would like to deepen it and, if possible, publish it within fiscal 2025.

Commentary · Kosei Torihiki no.891 pp.80–81

The January issue of Kosei Torihiki carries my column on hosting the first National Forum for Economic Law Research at Law Faculties — my impressions, an explanation of the forum’s purpose, and a plug for next year’s forum. Photos of the first forum are included too, so I would be glad if anyone curious about the atmosphere, or even slightly interested, took a look.

In fiscal 2025 it will be held at Kyoto University, in an even more upgraded form than the first!

Presentation

At the Committee for Improving the Status of Performers, I gave a presentation titled “Content Platforms and the Antimonopoly Act.”

With adult content distribution platforms in mind, and drawing on information the committee provided about the market environment and the characteristics of competitive conduct, I introduced and examined antimonopoly concerns and the basic ways of thinking about them.

As far as I know, there have been no antimonopoly violation cases in the adult content industry so far, and hearing about the contractual arrangements specific to the industry was very interesting. I expect that progress will come if the Japan Fair Trade Commission conducts a market study in the future.

In particular, the share of revenue taken by distribution platforms (as has come out in past court cases, it is apparently not unusual for the producer’s share to fall below 50%) is difficult to treat under antimonopoly law and looks likely to be an important subject of research.

The 6th meeting of the Committee for Improving the Status of Performers was held on December 3, 2024

Commentary · Japanese Yearbook of International Economic Law pp.247–251

I wrote a review of Eric Posner’s book for the Japanese Yearbook of International Economic Law.

This, too, is a theme I have studied since around 2019; in the end I produced only a draft and a table of contents and have not published a single article on it to this day. The main line of the book is the application of the Sherman Act to employers’ monopsony in labor markets. I have been thinking this problem is relevant in Japan too, but I have not found time to examine it properly. Meanwhile, study groups have been convened, a freelance act has been enacted, and the social situation has changed a great deal.