Google unveiled Gemini 3.5 and a suite of new AI capabilities at its I/O conference, including the Gemini Spark agent running on Google Cloud with multimodal tool support and Omni, which generates video from images, audio, and text. The company also introduced updates to products like Anti-Gravity 2.0 and Genie, its world-model-based navigation system powered by Street View and Waymo simulation data. Meanwhile, Elon Musk lost his high-profile lawsuit against OpenAI on statute-of-limitations grounds, ending his legal challenge to the company's for-profit transition. The week also saw Anthropic agree to a massive $30 billion funding round at a $900 billion valuation, signaling accelerating capital concentration in the AI industry, while Cerebras surged 90% in its IPO. OpenAI-Apple partnership tensions also surfaced during an otherwise dominant period for the ChatGPT maker. On the research front, OpenAI's models solved an 80-year-old Erdős geometry problem, and the field made progress on interpretability, agent benchmarks, and AI cyber capabilities. The week's additional developments ranged from deepfake enforcement efforts and new image-verification tools to escalating demonstrations of autonomous AI hacking and concerns about rapidly advancing AI security risks.