Insights

Latest Digital Insights

Practical insights on cybersecurity, AI, system integration, and digital transformation.

博物館 AI 用得比規則快：從 TTXC 近百萬人次與 UNESCO 全球調查，看文化場域的 AI 治理功課
Industry Trends
Updated: 2026.11.06

Museums Are Adopting AI Faster Than They Write the Rules: Lessons on AI Governance for Cultural Venues from TTXC and the UNESCO-ICOM Global Survey

The 2026 TTXC Taiwan Technology X Culture Expo drew 987,000 visits in 19 days, showing that the public has embraced AI interactive exhibits. Yet the UNESCO-ICOM global survey finds that while 57% of museums already use AI, 55% have no internal rules and only 17% train their staff. Drawing on Articles 4, 14 and 19 of Taiwan's AI Basic Act, this article outlines three risk areas (accuracy, copyright and visitor data) and five steps museums and curatorial teams can take now.
地端AI與開源模型的真實算式：token便宜了60倍，企業為何還是擴展不出去？
Tech Trends
Updated: 2026.09.23

The Real Math of On-Premise AI and Open Models: Tokens Got 60 Times Cheaper, So Why Can't Enterprises Scale?

Open models now carry a large share of real-world usage, inference prices have fallen roughly 60 times in four years, and more companies are weighing a move of AI back into their own data centers. Yet Mozilla's September report shows open models still reach production 12 points less often than closed ones, with the gap widening as companies grow. Surveys from Taiwan's MIC and of local SMEs point the same way: what blocks AI from scaling is not the model or the budget, but data integration and operational capability.
被打進來不可怕，回不來才可怕：從九月台灣三起資安事件，看企業該補的復原力與通報速度
Security Column
Updated: 2026.09.19

Getting Breached Is Not the Disaster, Not Coming Back Is: Resilience and Reporting Speed After Taiwan's September Incidents

September brought a run of intrusions at Taiwanese companies. Some had online services down for over a week, others completed containment, regulatory notification, and a market disclosure the same night. The difference was not the height of the perimeter but whether recovery and reporting were designed into the architecture. Meanwhile the EU Cyber Resilience Act's reporting duties took effect on 11 September, turning notification speed from crisis PR into legal liability.
AI 用得很勤，獲利卻沒動：2026 下半年企業 AI 落地的三道關卡
Industry Trends
Updated: 2026.09.06

Busy With AI, Flat on Profit: The Three Gates of Enterprise AI Adoption in Late 2026

McKinsey's latest survey finds 80% of respondents feel a personal productivity lift from AI, yet only 37% see any impact on company earnings. From converging agent interoperability standards to autonomous AI attacks on Taiwan and compliance clocks now running in both Europe and Taiwan, what decides AI adoption in late 2026 is process redesign, system integration, and governance.
AI重寫產品團隊的分工邏輯：為什麼「清理」與「維護」才是數位轉型的真本事
Tech Trends
Updated: 2026.07.06

AI Is Rewriting How Product Teams Divide Labor: Why "Cleanup" and "Maintenance" Are the Real Skills of Digital Transformation

Generative AI has made building easier than ever, moving the handoff from documents to working prototypes. Yet from our front-line experience implementing enterprise systems, the truly scarce roles that decide project success are rarely the builders, but those who know how to clean up and maintain for the long run.
機器人流量首度超越人類：AI 代理時代的企業資安新課題
Security Column
Updated: 2026.06.06

Bot Traffic Surpasses Humans for the First Time: New Security Challenges in the Age of AI Agents

Global web traffic has crossed a historic line — bots have surpassed humans for the first time, reaching 57.5%, a year earlier than predicted. The driver is not traditional crawlers but the fast-rising tide of "AI agent" traffic. As more requests come from automated agents, enterprises face new security challenges in traffic identification, access control, and content protection.