AI Update: National Security, Tariffs, and a $20B Startup Without a Product

Welcome to the newest edition of ID Tech’s AI update. Here’s the latest big news on the shifting landscape of AI and identity technology:

Australia has banned DeepSeek from all government devices and systems, with the government explaining that the China-made AI tool poses “unacceptable risk” to national security. Meanwhile, multiple South Korean government ministries, including the Ministry of National Defense, the Ministry of Science and ICT, and the Ministry of Foreign Affairs have also blocked access to DeepSeek on their computers.

The inaugural Director of the US AI Safety Institute has resigned. Elizabeth Kelly’s departure is widely seen as a political development stemming from the arrival of a new administration in the White House, and comes just ahead of a Paris AI summit that will feature various world leaders—but not technical staff from the US AI Safety Institute.

MediaTek is running business simulations in preparation for potential US-imposed tariffs. The Taiwan-based chip designer has been working with Nvidia on the GB10 Grace Blackwell chip for AI supercomputers, and has previously predicted that its AI Application-Specific Integrated Circuit (ASIC) business will generate over $1 billion in revenues by 2026.

Safe Superintelligence (SSI) is in talks to raise funds at a valuation of $20 billion. The startup was co-founded by Ilya Sutskever, who also co-founded OpenAI. It previously raised $1 billion from investors including Sequoia Capital and Andreessen Horowitz in September, at a valuation of merely $5 billion. SSI has not launched a product and does not have revenues.

Anthropic is in talks to raise over $2 billion in a new financing round that includes Menlo Ventures, Bessemer Venture Partners and Abu Dhabi’s MGX. The oversubscribed funding round would reportedly value Anthropic at over $60 billion. Anthropic is widely seen as a direct rival to OpenAI thanks to its beloved Claude chatbot.

France and the United Arab Emirates have agreed to build a 1 gigawatt AI data center in France. The agreement entails an investment somewhere between €30 to €50 billion.

Meanwhile, France-based AI startup Mistral has entered into a partnership with Cerebras Systems, an AI chip company backed by the UAE’s G42 group. Cerebras’ tech is powering Mistral’s new “Le Chat” AI chatbot, which is raising eyebrows with its 1,000-word-per-second response rate.

OpenAI has launched “Deep Research”, one of its first agent-style tools that can perform complex actions autonomously. In this case, the agent is designed to undertake thorough web research in order to write PhD-level reports that would normally take humans many hours. According to some academic types who have early access, it’s quite good.

The chatbot’s take: Time for an AI market vibe check:

–

February 7, 2025 – by Alex Perala