Issue #54

Made by Google '26, Anthropic watermarks, and a supply-chain breach

Thursday, August 13, 2026

Today covers Google's Pixel 11 and Gemini announcements, a backlash over Anthropic's new Claude watermarks, and fresh AI funding for Lovable, Cognition, and Thrive Holdings. We also look at Amazon's default use of Twitch content for training and a large credential leak.

1 of 7  ·  TechCrunch AI

Everything announced at Made by Google ’26: Pixel 11, Pixel Watch 5, Pixel Tag, and tons of Gemini features

Google unveiled the Pixel 11 series, Pixel Watch 5, and a new item tracker called Pixel Tag at its Made by Google 2026 event. The company also introduced a range of new Gemini AI features across its hardware lineup.

Read the full story →

2 of 7  ·  TechCrunch AI

Some Claude users are mad that Anthropic’s new watermarks will catch them using it at their jobs, classes

Anthropic has introduced a watermarking system for content generated by its Claude AI model, allowing outputs to be identified as AI-generated. Some users have complained on social media that the feature could expose their use of the tool in workplaces and classrooms.

Read the full story →

3 of 7  ·  TechCrunch AI

Lovable confirms new $13.3B valuation, raises another $400M

Lovable confirmed a new $13.3 billion valuation alongside a $400 million funding round. The raise follows the startup reaching $500 million in annualized run rate revenue in June.

Read the full story →

4 of 7  ·  TechCrunch AI

AI coding startup Cognition reportedly already in talks to raise at $40B valuation

Cognition, an AI coding startup, is reportedly in talks to raise a new funding round at a $40 billion valuation. The move comes just months after the company raised $1 billion at a $26 billion valuation.

Read the full story →

5 of 7  ·  TechCrunch AI

OpenAI-backed Thrive Holdings raises $2B to bring AI to the enterprise

Thrive Holdings raised $2 billion at a $12 billion valuation from investors including SoftBank, D1 Capital Partners, and Altimeter Capital. The OpenAI-backed company aims to bring AI to enterprise customers.

Read the full story →

6 of 7  ·  TechCrunch AI

Amazon will train on Twitch streamers’ content by default, unless they opt out

Twitch will use streamers' content to train Amazon AI models by default, requiring creators to opt out if they do not want their content included. Twitch CPO Mike Minton acknowledged that an opt-in system would see little participation, stating that few users would choose to opt in.

Read the full story →

7 of 7  ·  Ars Technica

Terabytes of credentials leaked in massive supply-chain attack

A supply-chain attack compromised an AI package, exposing terabytes of credential data. The breach affected 2,500 users whose data was scraped and exfiltrated.

Read the full story →

That's your scan for Thursday, August 13. Follow the links for the full stories.

Morning Roost Issue #54

Get this in your inbox every morning

Free. No spam. Unsubscribe anytime.

Subscribe free →