Latest / HotTea: The AI Morning Briefing / Malaysia weighs Huawei chips for a $494 million sovereign AI plan
Transcript
- Sections 00:00 What we're covering today 00:27 1. Malaysia weighs Huawei chips for a $494 million sovereign AI plan 02:21 2. OpenAI says agents now log 3.1 workdays for each human research day 03:51 3. China's top court made notice the trigger for AI liability 05:11 4. The U.N. rights chief asked countries to draw AI red lines together 06:33 5. Taiwan offered $20 billion more for U.S. chip investment while Washington kept the tariff threat 07:54 Visit Hot Tea Disclosure Narration uses an AI-generated voice. Transcript Welcome to Hot Tea for Tuesday, September 8, 2026. Malaysia weighs Huawei chips for a $494 million sovereign A I plan. Today's briefing covers the lead, models, politics, war, and companies. Bloomberg reports that Malaysia is seriously evaluating Huawei hardware for a two billion ringgit sovereign A I project. The report cites people familiar with the plan. Malaysia weighs Huawei chips for a $494 million sovereign A I plan. Bloomberg reports that Malaysia is seriously evaluating Huawei hardware for a two billion ringgit sovereign A I project. That's about four hundred ninety-four million U S. dollars. The report is based on unnamed people familiar with the plan, and says the system could use Huawei's Ascend 910C chips. State-linked Telekom Malaysia could take the main operating role. No choice is final. Bloomberg says the chip count is unclear, and Malaysia hasn't announced a purchase or supplier. The country is also discussing another tender that would give U S. suppliers a clearer chance to compete. The Prime Minister's Office, Malaysia's Ministry of Digital and Huawei did not respond to requests for comment. The project is meant to keep national data under Malaysian control, including government and security information. Prime Minister Anwar Ibrahim has said Malaysia doesn't want to remain only a buyer of foreign technology. His government has also raised concerns about the U S. Cloud Act, which lets U S. courts order American companies to provide data stored abroad. Bloomberg says Malaysian officials view the decision as commercial, while U S. officials frame the hardware as a security issue. Neither position proves a technical risk or a completed deal. A government announcement, tender document, purchase order or named chip count would make this a firm decision. Confirmation from Telekom Malaysia would also clarify who will run the system, while another tender could bring Nvidia or AMD into the competition. Washington's response matters too. A license warning, financing offer or diplomatic deal would show how hard the U S. pushes to keep allied A I systems on American hardware. Malaysia's decision will also show what sovereign A I means in practice. It could stop at local data control, or extend to choosing the whole technology stack. OpenAI says agents now log 3.1 workdays for each human research day. OpenAI published an internal snapshot of coding agent use across its research group. By mid-August, the company says total agent runtime equaled 3.1 eight-hour workdays for every human workday. At API prices, the median researcher was using more than six hundred dollars of inference each day. A user at the ninetieth percentile was using more than seven thousand dollars. OpenAI says researchers are writing code and running experiments faster. It now calls the system an automated research intern. That label covers well-defined jobs that can take a skilled researcher several days. These claims come from OpenAI's own internal data. Agent runtime isn't the same as research output, and OpenAI says its measurements are preliminary. Available compute also grew, while high-level planning remained a small share of agent output. More than half of successful four-to-eight-hour tasks still needed at least one human intervention. The company is targeting an automated A I researcher by March 2028, but that's a goal, not an observed result. The useful measure is accepted research work per dollar and per human review hour. Watch for outside replication of the task taxonomy, measured experiment quality and the rate at which agent work survives review. OpenAI should also report whether more agent runtime shortens model development, or simply moves the bottleneck to compute, safety review and research judgment. China's top court made notice the trigger for A I liability. China's Supreme People's Court issued the country's first judicial rules focused on A I disputes. Reports citing the court say an identifiable A I copy can violate personality rights. Those reports say the rule covers a person's face or voice when someone creates or shares the copy without consent. A service provider can face liability after receiving notice that its system generated infringing material and failing to act. Users can also be liable when they deliberately prompt a system to harm someone. The guidance covers doxxing, unfair algorithmic pricing, false output, model training and automated driving. Current English reports rely on the court's press conference, Xinhua and People's Daily rather than a complete official English text. China still has no single A I law, and the guidance leaves some disputed areas open until courts gain more experience. Lower courts will decide how notice, consent, reasonable action and proof work in real cases. The first cited judgments will show whether the rules change company behavior. Watch for injunctions against deepfakes, damages after ignored notices and cases that define what training use requires consent. Autonomous driving cases will also test how courts divide fault among a driver, a carmaker and an A I service provider. The U.N. rights chief asked countries to draw A I red lines together. Volker Türk, the U.N. High Commissioner for Human Rights, asked countries to agree on A I red lines. His call covered countries that host A I companies and countries in their supply chains. He also called for independent verification and stronger security work across the industry. Türk said he would write to A I companies in the coming days. He wants them to reduce the risks they can control now, and he tied the issue to jobs, democracy, environmental effects and system safety. This is a demand, not a binding rule. The speech names no threshold, inspection body, timetable or path for enforcement. Türk cited reported lab incidents to argue that safety systems trail model capability, but he offered no new technical assessment of those incidents. Governments also disagree on who should verify frontier systems and what companies must reveal. Türk's letters and the companies' replies will show whether this becomes a disclosure request or a wider safety plan. A Human Rights Council resolution or a member-state group could give the idea a formal path. But a red line still needs a measurable trigger, a named way to verify it and a consequence when a lab crosses it. Taiwan offered $20 billion more for U S. chip investment while Washington kept the tariff threat. At the SEMICON trade show, Taiwan presented its chip industry as a reliable partner for countries building A I systems. Economy Minister Kung Ming-hsin said Taiwanese companies plan another twenty billion dollars of U S. investment. TSMC is already investing two hundred sixty-five billion dollars in Arizona plants. The offer comes as U S. officials press more chipmakers to build inside the United States. They also warn that semiconductor tariffs are still coming for companies that don't build there. The twenty billion dollars is a plan spread across companies, not a disclosed set of signed projects. Moving production raises costs, and it can't quickly replace Taiwan's dense supplier network. Taiwan wants foreign support while keeping the industry that gives it economic and diplomatic weight. The United States and Europe want more production at home for the same reason. Company filings and site announcements will show how much of the planned investment becomes firm. U S. tariff terms and new European projects will matter too. The harder question is what happens to Taiwan's share of advanced production after those overseas plants open. Taiwan is trying to strengthen its alliances without weakening the industrial cluster at home. That is the signal before the noise. This briefing was produced from Hot Tea's verified daily edition. For the complete briefing and every source link, visit Hot Tea dot A I.