Claude Sonnet 5.5 Beats Opus as ChatGPT's $500 Plan Sparks Backlash

Anthropic's Claude Sonnet 5.5 scores 70% on Terminal-Bench at $2 per million input tokens, and OpenAI's new $500 ChatGPT plan has users calling it a rug pull. In today's AI and tech news roundup, I break down the White House AI safety accord, MI5's warning to UK universities, OpenAI Dev Day 2026 (GPT-6.1 Sol, Dots, ChatGPT Spaces, and Codex in the cloud), OpenAI's reported $30B raise at a $1.4T valuation, Broadcom's chip financing for Anthropic, and Google's Gemini 4 Argon. Chapters: 00:00 Intro 00:14 White House AI safety accord 01:12 MI5 warns UK universities about AI research 01:36 OpenAI Dev Day: GPT-6.1 Sol, Dots, Spaces 03:24 Codex moves fully to the cloud 03:55 Claude Sonnet 5.5 beats every benchmark 06:55 OpenAI eyes a $30B raise at $1.4T 09:34 ChatGPT's $500 plan backlash 11:21 Broadcom could finance Anthropic's chips 12:48 Google announces Gemini 4 Argon 15:25 Outro Sources: Al Jazeera: How does Trump's White House AI accord work? The Guardian: MI5 warns UK universities over theft of tech secrets OpenAI: DevDay 2026 Recap SiliconANGLE: OpenAI Codex cloud coverage Bloomberg: OpenAI Targets $30 Billion in Funding at $1.4 Trillion Value Business Insider: ChatGPT's new plan costs $500 a month CNBC / Reuters: Broadcom to lend Anthropic up to $42 billion to lease its chips Google: Gemini 4 Argon, our next era of frontier intelligence Some figures are reports or claims as cited in the video, not confirmed facts. Audio was cleaned with Adobe Podcast Enhance. Background music: "Dystopia" by Neutrin05, sourced from BreakingCopyright (royalty-free music). #AI #ArtificialIntelligence #TechNews