DeepSeek plans to develop its own AI chips to bypass US export controls
DeepSeek, the Chinese startup behind competitive LLMs, is planning to enter the silicon business by developing its own AI chips, according to Reuters. The company has been working on the move for about a year, meeting with potential hardware partners and hiring engineers for the project. This move is seen as a response to US export controls that restrict Chinese access to advanced semiconductors.
Entities
Related
Alibaba launches agentic AI tool for enterprises with Slack, Teams integration plans
Alibaba unveiled a new enterprise AI tool on July 11, 2026, designed as an agentic assistant for businesses. The tool is planned to integrate with Slack and Microsoft Teams, signaling Alibaba's push into enterprise productivity AI.
Apple sues OpenAI for allegedly stealing AI hardware trade secrets
Apple has filed a lawsuit against OpenAI, accusing the company of stealing AI hardware trade secrets in a scheme allegedly driven by OpenAI's head of hardware and other levels of the organization. The lawsuit highlights a pattern of alleged malicious practices by OpenAI.
Lidl owner plans to build EU-funded AI gigafactory
The owner of supermarket chain Lidl has announced plans to build one of several AI 'gigafactories' backed by the European Union. The facility would be part of a broader EU initiative to boost sovereign AI infrastructure. The news was reported by Cybernews on July 12, 2026.
Software engineer publishes final part of LLM-from-scratch series covering inference and decoding
A software engineer published the fourth and final part of a blog series explaining LLMs from the ground up, focusing on token-by-token generation, KV cache, and decoding strategies (temperature, top-k, top-p). The series aims to help other software engineers understand the internals of LLMs.
Microsoft Foundry adds Hugging Face models to managed compute platform
Microsoft Foundry, a platform for building agentic AI applications, now includes Hugging Face models alongside offerings from Microsoft, OpenAI, Anthropic, Meta, Mistral, and DeepSeek. The platform provides a single endpoint and SDKs in Python, C#, JavaScript, and Java, with managed compute for deploying and scaling models. This expands model selection for developers using Foundry's multi-agent orchestration and memory features.
