Three of Japan's Major IT Companies Announce Partnerships with Anthropic

In the span of just one month, NEC, Hitachi, and Fujitsu have each announced collaborations with Anthropic. Having all three companies partner with the same AI vendor is an unusual move, signaling that competition over the standardization of AI foundations among Japanese enterprises is heating up.

Fujitsu is drawing particular attention. Through a strategic partnership, the company has reportedly secured early access to **Claude Mythos Preview**, Anthropic's latest AI model. However, details remain unclear as to whether Fujitsu can actually put this latest model to use at this point, and further developments are awaited.

Each company appears to have a different aim: NEC is focused on strengthening its enterprise AI solutions, Hitachi on applying AI to social infrastructure, and Fujitsu on leveraging the technology for research and development as well as business transformation. In any case, there seems little doubt that Anthropic's safety-first approach is attractive to Japan's major corporations.

NVIDIA's Jensen Huang: "AI Layoffs Are a Lazy Excuse"

NVIDIA CEO Jensen Huang had harsh words for corporate executives who cut headcount on the grounds of AI adoption. According to a report by Business Insider, Huang said it is "lazy" for CEOs to use AI as an excuse for layoffs.

Opinions within the tech industry remain divided over AI's impact on employment. Some take an optimistic view, arguing that AI will create new jobs, while others raise the realistic concern that people are actually losing theirs. Huang's remarks come from his position at the top of an AI company, but for the workers affected on the ground, the message may be a painful one to hear.

US Law Enforcement Warns of "Anti-Tech Extremism"

According to a report by Ars Technica, US law enforcement agencies are growing more vigilant about "anti-tech extremism." As hatred and backlash toward AI mount, there are concerns about aggressive actions targeting technology companies and AI facilities.

As AI becomes more deeply embedded in society, backlash against the technology is emerging as a social phenomenon in its own right. Alongside regulatory responses, the importance of dialogue across society is only growing.

CUDA 13.3 Released — Drawing Attention in the Local LLM Community

A hot topic in Reddit's LocalLLaMA community is the release of NVIDIA's **CUDA 13.3**. The developer download page and release notes have been published, and the community has already begun testing compatibility with llama.cpp.

CUDA updates directly affect the speed and stability of local LLMs, which makes them significant for the open-source AI community. Improvements in inference performance on consumer GPUs are particularly anticipated.

IBM Granite 4.1-30B Overshadowed by Qwen and Gemma?

IBM's **Granite 4.1-30B** model has been released, but some on Reddit point out that it has drawn little attention, overshadowed by the presence of Qwen 3.6 and Gemma 4. The fact that it is a dense model without reasoning capabilities is cited as one reason, though it may still be worth evaluating for specific use cases such as coding and RAG.

The IBM Granite team has commented that models with reasoning capabilities are also in the works, and expectations are building for the next iteration.