AI News 5/14 Several noteworthy developments in the past 24 hours: DeepMind spin off company Isomorphic Labs has completed a Series B funding of $2.1 billion, led by Thrive Capital. The commercialization of AlphaFold is advancing faster than expected, and the distance from the laboratory to the industry for AI pharmaceuticals is shortening. OpenAI has opened up the GPT-5.5-Cyber security model to EU cybersecurity teams. EU institutions, member state governments, and some businesses will be the first to connect. The significance of this move is not just about launching the product - Anthropic's Claude Mythos has not yet announced a timetable for the EU landing, and OpenAI has taken the lead in government level AI security services. 3. Google has started a large-scale recruitment of "forward deployed engineers", a group of hundreds of people, dedicated to helping enterprise clients deploy AI products. This model is not new, Palantir has been using it for a long time, but Google's personal involvement in doing this shows that there is still a huge gap between their judgment of "selling APIs" and "helping customers use them" - and they choose to fill it themselves. Google I/O 2026 confirmation will be held on May 19th, 5 days from now. Confirmed highlights: preview of Android XR glasses (with hardware partners such as Samsung and Warby Parker in place), exposure of leaked images of Aluminum OS (Google's PC operating system aimed at replacing Chrome OS), and expected release of a new Gemini model. For ordinary users, the next two weeks are worth paying attention to the pace of new I/O product launches; For the industry, the competition among top companies in cybersecurity and government services is no longer just a trailer, but an ongoing drama.