AI news daily: what changed on October 7, 2026

This digest is assembled automatically every day: headlines are pulled from public AI news feeds, screened for duplicates, rewritten in original wording, and linked back to the source. Nothing here is a substitute for reading the original report.

1. OpenAI drops another batch of mathematical breakthroughs

OpenAI has released 722 manuscripts with solutions to long-standing math problems. These manuscripts cover 372 result families and were produced by an unreleased model. The release extends a series of math breakthroughs from OpenAI.

Source: The Verge AI

2. Ex-Ramp engineers raise $20M for platform Melius after scrapping their first product

Melius has secured $20M in funding. The company was founded by former Ramp engineers. It is now developing tools to create campaign assets.

Source: TechCrunch AI

3. How AI decision models could change content moderation

Musubi introduced PolicyLM-1.7B, a decision model for real-time content moderation. The model is lightweight and has open weights. It was announced by Musubi on Tuesday.

Source: TechCrunch AI

4. AI computing startup Lambda to raise $4B ahead of planned IPO

AI computing startup Lambda to raise $4B ahead of planned IPO was reported by TechCrunch AI. See the original report for details.

Source: TechCrunch AI

5. EmbeddingGemma 2: an open, lightweight multimodal embedding model

Google DeepMind Blog introduced EmbeddingGemma 2, a multimodal embedding model. It is open and lightweight. EmbeddingGemma 2 was announced on the Google DeepMind Blog.

Source: Google DeepMind Blog

6. The next hurdle for AI agents: getting websites to let them in

Personal AI agents are designed to perform tasks like shopping and booking flights. However, some websites block or limit these agents using anti-bot defenses. A new standard is being developed to address this issue.

Source: TechCrunch AI

7. Hark releases an AI personal assistant with a focus on privacy

Hark has released a personal assistant that uses AI technology and prioritizes user privacy. This operating system is intended to be a competitor to existing products like Muse, Dots, and Instinct. It is developed by Hark's AI lab.

Source: TechCrunch AI

8. Mirror Particle is building a 'world model' of human behavior

Mirror Particle is developing a model to predict human behavior. The company will launch at TechCrunch Disrupt's Startup Battlefield 200. It aims to improve market research and brand strategy by addressing limitations of large language models.

Source: TechCrunch AI

9. We can't just change the definition of 'recording'

Tech companies are reexamining what constitutes a recording with new AI hardware. Devices with microphones or cameras were typically seen as either recording or not. Microphones and cameras capture sound and images respectively.

Source: The Verge AI

10. Anthropic is giving startups a free year of Claude Team and $1,000 in credits

Anthropic has a program for startups that includes a free year of Claude Team and $1000 in credits. The company created this program to support businesses building on top of AI models. Anthropic aims to help people benefit from AI through the companies that use these models.

Source: TechCrunch AI