President Donald Trump issued an executive order on June 3, 2026, targeting cybersecurity risks associated with artificial intelligence. The order mandates the prioritization of cyber defense and the creation of a voluntary clearinghouse to help patch software vulnerabilities. It also requests a 30-day review period for new models from AI companies. This policy is notably less stringent than an earlier draft, which proposed a 90-day voluntary review and had been reviewed by OpenAI, Anthropic, and Google. Administration officials reportedly scaled back the order following warnings that stricter regulations could stifle innovation.

In financial news, the Chinese AI startup DeepSeek is nearing a funding round valued between $52 billion and $59 billion. The company aims to raise approximately 50 billion yuan ($7.4 billion) from investors including Tencent and CATL, with founder Liang Wenfeng contributing 20 billion yuan of his own capital. DeepSeek previously gained attention for developing low-cost V3 and R1 models that challenged the industry expectation that high-end AI research requires massive capital expenditure.

Microsoft unveiled its new reasoning-focused model, MAI-Thinking-1, during its Build developers’ conference. The company describes the product as a medium-sized model trained on commercially licensed, clean data without distillation from third-party sources. Microsoft reported that the model performs competitively against industry benchmarks and exceeds Anthropic’s Sonnet 4.6 in blind anecdotal tests. MAI-Thinking-1 is currently available in private preview via Microsoft Foundry and is designed for enterprise applications including scientific reasoning and complex multi-layered instruction tasks.

President Donald Trump issued an executive order on June 3, 2026, targeting cybersecurity risks associated with artificial intelligence. The order mandates the prioritization of cyber defense and the creation of a voluntary clearinghouse to help patch software vulnerabilities. It also requests a 30-day review period for new models from AI companies. This policy is notably less stringent than an earlier draft, which proposed a 90-day voluntary review and had been reviewed by OpenAI, Anthropic, and Google. Administration officials reportedly scaled back the order following warnings that stricter regulations could stifle innovation.

In financial news, the Chinese AI startup DeepSeek is nearing a funding round valued between $52 billion and $59 billion. The company aims to raise approximately 50 billion yuan ($7.4 billion) from investors including Tencent and CATL, with founder Liang Wenfeng contributing 20 billion yuan of his own capital. DeepSeek previously gained attention for developing low-cost V3 and R1 models that challenged the industry expectation that high-end AI research requires massive capital expenditure.

Microsoft unveiled its new reasoning-focused model, MAI-Thinking-1, during its Build developers’ conference. The company describes the product as a medium-sized model trained on commercially licensed, clean data without distillation from third-party sources. Microsoft reported that the model performs competitively against industry benchmarks and exceeds Anthropic’s Sonnet 4.6 in blind anecdotal tests. MAI-Thinking-1 is currently available in private preview via Microsoft Foundry and is designed for enterprise applications including scientific reasoning and complex multi-layered instruction tasks.