Provenance, safety, and AI for science: the responsibility front

Four articles held together by one question: who vouches for what AI produces, and what do we use it for that's worth doing? C2PA Content Credentials enters Gemini, Search, and Chrome. SynthID, Google's watermark for generated content, is adopted for the first time by OpenAI, ElevenLabs, and Kakao: the major labs converge for the first time on a shared provenance mechanism. A story largely missed by mainstream media, structural for the media literacy of the next decade. On the applied front: Gemini for Science is the multi-agent platform for scientific research; WeatherNext, with the case of hurricane Melissa predicted five days ahead, shows what a specialized, well-grounded model can do. It's the least spectacular and most interesting front for those who think of AI as a collective tool, not a gadget.
Google DeepMind's AI weather model predicted Hurricane Melissa's rapid intensification to Category 5 with record lead time. A real-world case that resets the benchmark for early warning forecasting.
For the first time the major AI labs converge on a shared provenance mechanism. SynthID, embedded inside Google for three years, leaves Mountain View and becomes an industry standard.
Hypothesis Generation, Computational Discovery and Literature Insights. DeepMind, Research, Cloud and Labs assemble a research toolkit, wired into 30+ life-science databases through Antigravity.
Is an image authentic or AI-modified? Google adopts the C2PA standard to give a verifiable answer, first in the Gemini app and then in Search and Chrome over the coming months.
Shopping becomes an agentic service
Workspace becomes an automation factory
The personal assistant stops answering and starts acting
Building software stops resembling writing and starts resembling directing
The web gears up to be read by agents
Nine announcements that redesign Google's physical ecosystem
Models, chips, and business models: Google chooses where to make margin
Content creation becomes modular