My experience with AI in publishing has focused on one central question: How should publishing organizations redesign work to create greater value? I see AI not as a standalone technology but as part of a broader transformation combining workflow redesign, knowledge management, and organizational capability development.

Three initiatives illustrate this in practice. First, an AI-assisted proofreading system for Thai-language publications, working alongside human editors, has reduced average errors per title by 67%; 132 titles are scheduled to run through it this year, with a projected saving of 31.2 hours of repeat proofreading. Second, AI support in publishing production has cut redundant tasks and process steps by more than 30%. Third, I built an internal sales-reporting web application that operates independently of our ERP, handling varied contract structures and validating data before migration to the production system. An AI-powered internal chatbot—now in development—will extend this foundation, targeting on-demand access to 100% of our product and title information. I have also led development of a digital rights management system for more accurate, efficient administration of contracts and licensing.

Across these projects, I have learned that successful AI adoption depends less on selecting the most advanced model and more on knowing precisely where each workflow creates value. My approach is to simplify, standardize, and improve data quality before automating: AI should strengthen efficient processes, not automate inefficient ones.

These results reflect my broader conviction: the opportunity is not to replace human expertise but to elevate it. As routine cognitive work shifts to intelligent systems, publishing professionals gain time for editorial judgment, relationship building, and intellectual property development: the work that only people can do.

Thailand’s conversation about AI and publishing has moved quickly from experimentation to governance. In July 2026, the Electronic Transactions Development Agency (ETDA) released a revised draft of Thailand’s first AI Act for public consultation—a risk-based framework that would impose transparency obligations for AI-generated content, with mandatory labeling under consideration. Most relevant to publishers, ETDA has convened a committee of intellectual property specialists, legal experts, and judges to define acceptable boundaries for training-data usage and revenue-sharing, responding to the unresolved dispute over AI models trained on copyrighted works without permission.

The market behind this debate is growing again. The Thai book industry recovered from roughly 12.5 billion baht in 2021 to an estimated 18 billion baht in 2025, and the Publishers and Booksellers Association of Thailand (PUBAT) projects 5–8% growth in 2026. This year’s Bangkok International Book Fair drew over 1.3 million visitors—with Generation Z making up around 70% of attendees at recent national book events—and the Bangkok Rights Fair produced more than 600 rights negotiations, notably with Korean, Taiwanese, and Chinese partners. PUBAT is also championing a “book economy” strategy that extends Thai stories into film, series, games, and webtoons.

Against this backdrop, I see the main opportunity in using AI to remove operational friction—metadata, production tracking, reporting, knowledge retrieval—so that a growing market can be served by leaner publishing operations. The main risks are unresolved copyright ownership of AI-trained and AI-generated works, and a widening capability gap between large publishing groups and smaller independent presses. Publishers who engage in the governance debate now, rather than after the law settles, will help shape rules the whole industry can work with.