Urgent.News

What's breaking now, across thousands of outlets.

AI

Apple trains its own AI model for China market with Alibaba’s support

Apple trains its own AI model for China market with Alibaba’s support

Apple has developed its own large language model tailored for the Chinese market, collaborating with Alibaba Group, according to sources familiar with the matter. This marks a shift from Apple's previous reliance on third-party AI models to power features on its devices in China, where American AI models like Anthropic's Claude and OpenAI's ChatGPT are unavailable.

The partnership, not previously reported, has not been confirmed by Apple or Alibaba. Apple Intelligence, the company's suite of AI tools, is set to launch in China in the coming months, following an update to its iOS operating system. This move signals Apple's desire to gain greater control over the AI experience in the competitive Chinese market, where it has fallen behind local rivals like Huawei.

The collaboration with Alibaba and potential involvement from Baidu marks a unique dual-track strategy for Apple in AI deployment in China. The Cyberspace Administration of China recently registered Apple's generative AI service, clearing the regulatory hurdles that had previously limited access for many U.S. technology firms.

Written by urgent.news from The Indian Express's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at indianexpress.com →

More in AI

Dead Auto-Skills Were Padding Every Conversation: A Weekly Curator That Flags at 30 Days and Archives at 90

"Claude Code gets smarter the more you use it" is only half the story. The other half is that it gets heavier. Six months ago I was a university student making ¥100k a month.

  • Claude Code's abilities improve with usage but increase monthly token bill
  • Skills accumulate, becoming unused and burdensome over time
  • Weekly automated curation pass detects and archives stale skills

AI/ML Research Digest — Aug 02, 2026

Reliable provenance and graded trust Explicit reliability modeling cuts hallucination. Σ‑Mem stores symmetric competence states for peer agents, while LedgerMind adds a Structured Evidence Ledger that…

Elon Musk makes a promise to employees at townhall

SpaceX aims to enhance its AI model Grok by integrating insights from its workforce. Elon Musk envisions this initiative as a means to embed human values deeply into the AI's framework. Employees are not only invited to enhance the internal AI systems but are also inspired to share their expertise.

How do you unit test an agent skill?

Agent skills are prompts, not code, and there’s no compiler to catch a broken one. Agent skills ship on the honour system. You rewrite one, run it twice, post something convincing in Slack, and that’s the review. Is it faster? More reliable? Going to cost more? This isn’t a strategy. This is astrology for prompts. That bugged me.

  • Agent skills are prompts reviewed on a trust system, not code.
  • skilleval tool runs a real agent with SKILL.md file, prompt, and fixture.
  • Tests assert results against criteria like cost, activated skills, and final message.

More from Sunday 16 August →