Microsoft sells OpenAI models in China. OpenAI and Anthropic won’t.

artificialintelligence-news.com Jun 18, 2026

Microsoft sells OpenAI models in China. OpenAI and Anthropic won’t.

Curated by Shubham · TechCapsules · Jun 18, 2026

AI-summarised brief · reviewed before publication

Microsoft has become the primary supplier of OpenAI models in China, selling the technology to the country's largest internet companies, including ByteDance, Ant Group, Meituan, and Tencent. The arrangement hands Microsoft a unique position, as it sells OpenAI models to Chinese firms that the model's creator, OpenAI, will not deal with directly. Microsoft's Azure AI revenue in China has expanded significantly, tripling in the financial year to June 2025. The company's contract with OpenAI allows it to set its own terms for selling GPT models abroad.

💡 Why It Matters

· Microsoft's dominance in the Chinese market for OpenAI models raises concerns about the potential misuse of the technology and the company's ability to police its use.
· The arrangement also highlights the tension between Microsoft's business interests and OpenAI's concerns about intellectual property and risk in the Chinese market.

Read full article on artificialintelligence-news.com

‘This will unveil the ground truth’: Why the Anthropic ban could benefit…

The US government's order to suspend access to Anthropic's best-performing AI models for non-US nationals has sparked concern in Europe, but some see it as an opportunity for the region. Uljan Sharka, founder and CEO of Italian AI model maker Domyn, views the move as a "huge opportunity" for Europe. The ban may unveil the ground truth and prompt European companies to develop their own AI models. This could lead to increased innovation and investment in the European AI sector, potentially benefiting companies like Domyn. The move has significant implications for the future of AI development in Europe, with some experts predicting a surge in domestic AI innovation.

💡 European companies may now focus on developing their own AI models, driving innovation and investment in the sector. This could lead to the creation of more tailored and region-specific AI solutions.

techcrunch.com · Jun 17

Anthropic’s latest feud with the Trump admin may actually help it, sales…

Anthropic surpassed OpenAI in market share of business spending for AI in May, according to Ramp. The AI lab raised $65 billion at a $965 billion valuation and filed confidential paperwork for an IPO after a profitable quarter. However, the Trump administration sent a letter demanding Anthropic ban non-Americans from accessing its models, forcing the company to pull its latest model from the market. Despite this, Anthropic's sales to businesses have not been deterred, with its share of AI subscriptions rising to 41% in May. The company's models, particularly Opus, remain popular among businesses, with Anthropic releasing a new version, Opus 4.8, in late May.

💡 The Trump administration's actions may boost Anthropic's reputation as a provider of powerful AI models, with its "dangerous" model generating significant interest. Anthropic's ability to attract businesses despite government controversy highlights its strong position in the market.

singaporelawwatch.sg · Jun 16

Is Anthropic hallucinating when it calls for an AI development break? –…

Artificial intelligence lab Anthropic has sounded an urgent alarm, warning that AI has begun building itself and humans are in danger of losing oversight. The company has called for a global pause on AI development, citing the risk of a minor error causing the technology to spiral out of human control. Anthropic released a 10,000-word report, When AI Builds Itself, highlighting the alarming rate at which autonomous systems are reducing human safety reviews. The report points to the capabilities of Anthropic's AI tool Claude, which has achieved significant increases in productivity and performance.

💡 The call for an AI pause highlights the growing disconnect between AI's rapid progress and the lack of effective guardrails. As AI leaders, the US and China must take responsibility for establishing safety protocols to prevent uncontrollable development risks.

techcentral.co.za · Jun 16

US scored ‘own goal’ with ban on top Anthropic model

The US government has banned top AI model Fable 5 and Mythos 5 from foreign nationals, citing national security concerns. Cybersecurity leaders at major US firms, including Nvidia and Adobe, have asked the Trump administration to lift the restrictions, arguing that they hinder efforts to prevent digital attacks. The ban will limit the industry's ability to find and fix software flaws, as rival models, including China's Kimi 2.7, offer similar capabilities. Anthropic has released a public version of Fable with cybersecurity safeguards, but the government believes a potential "jailbreak" could bypass these protections.

💡 The ban on Anthropic's AI models could create a cybersecurity gap, as China's open-source models are rapidly closing the gap with the best American ones. This overreaction by the government may inadvertently give Beijing an advantage in the AI arms race, as they likely have access to capabilities beyond what is publicly known.

asiatimes.com · Jun 16

Why the US government shut down Anthropic’s latest Claude AI model

The US government has suspended access to Anthropic's latest AI models, Fable 5 and Mythos 5, in response to an "export control directive" prohibiting use by non-US nationals. The directive is believed to be linked to a jailbreak discovered in Fable 5, which could bypass its safeguards and allow for nefarious purposes. Anthropic has been in conflict with the Trump administration over AI regulation and export policy, and the company claims the government's concern is based on research produced by Amazon engineers.

💡 The government's move highlights the tension between AI developers and regulators over the potential risks and benefits of advanced AI models. The incident also underscores the ongoing cat-and-mouse game between AI researchers and "jailbreakers" seeking to circumvent guardrails and exploit AI capabilities for malicious purposes.

artificialintelligence-news.com · Jun 15

The AI off switch: How Anthropic’s export controls sparked a global AI…

The US government issued an export control directive to suspend access to Anthropic's AI models, Fable 5 and Mythos 5, citing national security concerns. The directive, which took effect on June 13, 2026, requires Anthropic to disable access for all customers, including foreign national employees. The move was sparked by a method for "jailbreaking" Fable 5, which Anthropic disputed as a limited capability. The US government's action has set off alarm bells abroad, with Europe and Canada expressing concerns about AI sovereignty.

💡 The US government's move highlights the growing tension between AI safety and national security, with Anthropic caught in the middle. As the AI industry continues to expand, the question of who controls the AI that powers the world's systems is becoming increasingly pressing, with the US government's actions suggesting a shift towards greater regulation and control.