Issue #89
Anthropic's biology lab, Google's household agent, and a military AI scare
Saturday, September 19, 2026
Today covers Anthropic's new biology research lab, agentic AI tools from Google and Meta, and a novel model from a ChatGPT co-creator. We also look at an AI hallucination that nearly triggered a US military operation, Manus's $500M raise, and a $100M round for a physical AI startup studio.
1 of 7 · TechCrunch AI
Anthropic is operating a lab that conducts biology experiments
Anthropic has established a lab to conduct biology experiments, aligning with claims from AI leaders that the technology can accelerate disease research. The move comes as the company continues to warn about potential risks posed by advanced AI systems.
Read the full story →2 of 7 · TechCrunch AI
Google’s new ‘CC’ is an AI agent that helps families run their households
Google has reoriented its CC AI agent toward household coordination, allowing families to share emails, schedules, and tasks. The agent can manage calendars, fill out forms, create shopping lists, and plan meals.
Read the full story →3 of 7 · TechCrunch AI
Meta’s Muse hits Mac, letting the AI take actions on your computer
Meta has released Muse for Mac, an AI tool that can access users' files and applications to perform actions on their behalf. The expansion brings the agent-based assistant to Apple's desktop platform.
Read the full story →4 of 7 · TechCrunch AI
A new kind of AI model from a ChatGPT inventor is thrilling developers
Jev, a new AI model created by one of the developers behind ChatGPT, has launched with a focus on lower cost and faster performance for software applications. The model is drawing early attention from developers evaluating its efficiency.
Read the full story →5 of 7 · TechCrunch AI
AI hallucination nearly triggers US military operation
I don't have enough information from the provided content to write an accurate summary. The excerpt only includes a single quote from a GovAI research scholar about the uncertainty inherent to large language models, but it does not contain the details needed to summarize what actually happened. To write a factual summary, I would need the substantive content of the article, such as what the AI hallucination was, how it nearly triggered a military operation, and the outcome. If you can provide the full article text, I can produce a summary that meets the digest's standards.
Read the full story →6 of 7 · TechCrunch AI
Manus seeks $4B valuation in new $500M fundraise as it resumes independent ops
Manus is in talks to raise $500 million at a $4 billion valuation as it resumes independent operations. The company had earlier ended a planned merger with Meta.
Read the full story →7 of 7 · TechCrunch AI
A startup that builds other startups raised $100M and is all-in on physical AI
Vantora, formerly known as UP.Labs, raised $100 million to build startups focused on physical AI for industrial corporations.
Read the full story →That is today's roundup. We will be back tomorrow morning with the next digest.
Morning Roost Issue #89