Full Summary
This Saturday morning, Apple confirms it has developed its own large language model specifically for the Chinese market, a pivotal move corroborated by Reuters, the BBC, and Irish Sun. This China-specific model, created with support from Alibaba and expected to integrate Baidu's technology for search functions, marks a significant shift from Apple's previous reliance on external AI partners in the region. Both foreignpolicyjournal.com and irishsun.com highlight this allows Apple to gain more control over its AI experience in China and navigate strict local regulations. AzerNews and ETV Bharat emphasize this partnership is crucial for Apple to compete against domestic manufacturers and revive falling iPhone sales in the country. TechJuice and Decrypt confirm Apple chose Alibaba last year after discussions with several companies. The Cyberspace Administration of China approved Apple's generative AI service last month, clearing the way for Apple Intelligence to launch in China through an iOS update in the coming months. This makes Apple the first foreign company approved by Beijing to offer a proprietary AI model in mainland China. Meanwhile, new research from Moonshot AI, reported by the-decoder.com, shows that even leading multimodal AI models, including GPT-5.6 Sol, still struggle significantly with visual perception, with no frontier model reaching 60 percent accuracy on the new PerceptionBench test. In other AI news, Zhipu's GLM-5.3 AI model has outperformed Anthropic's Mythos 5 in a cybersecurity test, scoring 84.5% on CyberGym for identifying security flaws, as reported by NewsBytes and iNews Zoombangla. Zhipu plans to release GLM-5.3 publicly in about two weeks, offering advanced cyber-defense tools to developers. This means that while AI capabilities are rapidly expanding, with new, localized models like Apple's entering major markets, fundamental challenges in areas like visual perception persist.