AI News — Tuesday, June 2, 2026
Florida has filed a first-of-its-kind lawsuit against OpenAI and CEO Sam Altman, alleging their AI models contributed to violent incidents.
Alphabet is reportedly seeking to raise $80 billion to fund its ambitious AI infrastructure expansion, signaling massive investment in the sector.
Nvidia is expanding its reach into the CPU market, partnering with Microsoft, Dell, and HP to power new AI agent-enabled personal computers.
OpenAI has announced the availability of its frontier models and Codex on Amazon Web Services, expanding access for developers and enterprises.
OpenAI has published its stance on AI policy and political advocacy, outlining its approach to shaping future regulations and public discourse.
Researchers introduce GrepSeek, a novel method for training AI search agents to directly interact with and query large text corpora more effectively.
A new paper presents COLLEAGUE.SKILL, a system for automatically generating AI skills by distilling knowledge from human experts.
OpenAI is investing in building new data center infrastructure in Michigan, positioning it as a key hub for the 'Intelligence Age'.
This research explores Trust-Region Behavior Blending, a technique for improving on-policy distillation by combining expert and student policies.
An anecdotal account details a company's $660K AI platform purchase leading to job replacement and a critical system rollback, illustrating real-world AI implementation pitfalls.
Google shares insights into how its Gemini AI model was leveraged internally to develop and streamline the production of Google I/O 2026.
SwanVoice introduces a new method for highly expressive, long-form, zero-shot speech synthesis capable of generating both monologues and dialogues.
The technical report for Mellum2 provides a detailed overview of the model's architecture, training, and performance benchmarks.
This article discusses the evolving needs of non-technical builders in the AI era, emphasizing the shift from intuitive 'vibe coding' to more structured, clear thinking.
An article explores practical methods for constraining Large Language Models to behave more predictably and align with user expectations, similar to how human users are constrained.