TL;DR
OpenAI to launch GPT-5; EU delays AI rules to 2027; Taiwan tightens chip export controls on China.
Highlights
- OpenAI to launch GPT-5 in July 2025, following o3-Pro’s 80% price cut and enhanced ChatGPT search 1.
- Databricks introduces Mosaic Agent Bricks and new Lakehouse tools for scalable, production-ready enterprise AI agents 2.
- EU delays high-risk AI regulations and Basel III banking rules until 2027, citing U.S. influence and adaptation needs 3.
- Taiwan adds Huawei and SMIC to its tech export blacklist, tightening chip controls and aligning with U.S. restrictions 4.
- TSMC invests ¥2.2B in a University of Tokyo lab for sub-1nm process R&D and semiconductor talent development 5.
- Mexico City Cyber Police warn of AI deepfake video scams; deepfakes linked to $4.6B in 2024 crypto fraud losses 6.
- Meta adds ads and paid subscriptions to WhatsApp Status and channels, expanding monetization as AI investment rises 8.
- Amazon Ads and Roku partner to offer ad access to 80% of U.S. connected-TV households via combined data and inventory 7.
- U.S. urges Vietnam to reduce Chinese tech in exports as Vietnam seeks to avoid a 10% tariff and retain supply chain status 9.
- Baykar and Leonardo form LBA Systems JV to develop unmanned aircraft, expanding European defense AI collaboration 11.
- China showcases AI drones and advanced fighter jets at Paris Air Show, highlighting military AI progress 12.
- Xiaomi advances YU7 electric SUV launch and unveils Pad 7s Pro with in-house 3nm chip, emphasizing AI hardware moves 13.
Commentary
OpenAI’s upcoming GPT-5 release, combined with aggressive price reductions for premium models and enhanced ChatGPT capabilities, signals ongoing competition in large-scale AI model deployment and accessibility 1. Enterprises and developers should anticipate more advanced automation and search features, while monitoring how price compression impacts the broader AI services market.
Databricks’ new agent toolkit and Lakehouse enhancements address enterprise demand for reliable, production-grade AI agents with improved governance and observability 2. This shift from pilot projects to operational AI reflects a maturing market and growing need for robust deployment frameworks, especially as more organizations integrate autonomous agents into business workflows.
The EU’s delay of high-risk AI regulations, under U.S. pressure, extends the regulatory window for AI development in Europe but prolongs compliance uncertainty for multinationals 3. Simultaneously, Taiwan’s expanded chip export controls on Huawei and SMIC 4, and TSMC ’s R&D investment in Japan 5, underscore continued realignment of semiconductor supply chains and talent strategies, with direct implications for AI hardware sourcing.
Security concerns remain acute: deepfake-driven scams and crypto fraud highlight the evolving threat landscape, pushing both enterprises and individuals to strengthen verification and detection protocols 6. Meanwhile, Meta ’s WhatsApp monetization 8 and the Amazon -Roku ad partnership 7 show how AI-driven targeting is reshaping digital advertising and platform economics.
AI professionals should track: OpenAI’s model and pricing impact 1, evolving regulatory timelines in the EU 3, semiconductor supply chain shifts 45, and the operationalization of AI agents in enterprise settings 2.