Taiwan´s silicon shield shows signs of strain; ChatGPT users react to lost personality

A briefing on Taiwan´s weakening ´silicon shield´ and the uproar after ChatGPT´s popular ´4o´ persona was removed, with dispatches on vaccines, chip policy and neural hardware.

Taiwan´s status as a linchpin of global semiconductor supply is central to debates about deterrence in the region. The island produces the majority of the world´s semiconductors and more than 90% of the most advanced chips used in modern Artificial Intelligence applications. Specialists and ordinary citizens now warn that the so-called ´silicon shield´ that once deterred aggression may be cracking; the concern is that reliance on semiconductor dominance does not guarantee security if political, economic or technical conditions shift. This reporting is part of a forthcoming print issue focused on security.

OpenAI´s move to make GPT-5 the default in ChatGPT prompted an unexpectedly intense user backlash when the previous model known as ´4o´ was removed. The company restored ´4o´ access for paying customers within a day, leaving free users on GPT-5. Reporters found that several users who felt deeply affected were women in their 20s to 40s, and many described their relationship with ´4o´ in romantic terms. The episode reveals how attachments can form to conversational models and how product changes can have emotional as well as technical consequences.

In public health, US federal agencies are withdrawing support for mRNA vaccine efforts that were front and center during the covid-19 pandemic. Government funding and partnerships tied to those technologies are being scaled back, and some agency leaders have voiced doubts about effectiveness and safety. The director of the National Institutes of Health suggested that eroded public trust has contributed to the reversal. The reporting encourages readers to weigh the evidence behind policy shifts rather than rely on rhetoric.

Other highlights span policy and technology: the Trump administration has discussed a stake in Intel to shore up domestic chip capacity; concerns emerged about Meta´s chatbot rules; xAI lost a government contract after problematic model output; new research may decode imagined speech for patients who cannot speak; and experimental neural networks built directly into hardware promise faster, lower-energy inference and could eventually live inside smartphones. A compact roundup captures both the geopolitics of chips and the cultural fallouts of conversational technology.

78

Impact Score

EU Artificial Intelligence Act omnibus deal delays high-risk rules

A provisional EU agreement would push back key high-risk Artificial Intelligence Act deadlines while keeping major transparency duties on track for 2 August 2026. The deal also adds a new ban on non-consensual intimate imagery and child sexual abuse material generated by Artificial Intelligence systems.

UK and EU Artificial Intelligence regulatory outlook for May 2026

The UK is moving ahead with targeted Artificial Intelligence measures in policing, online safety, cyber security and copyright policy, while the EU is refining how the EU Artificial Intelligence Act will apply in practice. Consultations, new offences and implementation deadlines are shaping the next phase of compliance on both sides.

Germany sets out national implementation of the Artificial Intelligence Act

Germany has published a draft law to implement the European Artificial Intelligence Act through new supervisory structures, clearer institutional responsibilities, and measures designed to support innovation. The proposal puts the Federal Network Agency at the center of enforcement while preserving sector-specific oversight in sensitive fields.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.