AI News Today — Model Releases, Benchmarks & Industry Moves

GPT-5.5 Tops the List at 94.8: 2026-11-04 Smoke Quick Test Data Brief

The 2026-11-04 YZ Index Smoke Quick Test covered 12 models, with GPT-5.5 ranking first for the day at 94.8. The daily 10-question test is a short-term signal and is not equivalent to the weekly Full ranking.

TC

Google will let schools disable its AI tools, but not without sacrificing Search

Schools choosing to restrict AI altogether will no longer be able to access Google Search, as AI features are now integrated into the search engine.

TC

Anthropic is still grappling with the economics of compute

At a moment when Anthropic's compute margins should be getting better, they’re getting worse.

ARS

Google will let schools skip AI tools, but they have to give up search, too

Google will offer schools AI tools that purportedly don't give away the answers.

TC

The TechCrunch Founder Summit agenda is only half the reason to be in Boston on Nov. 4

On Nov. 4, TechCrunch Founder Summit 2026 brings founders and VCs to Boston's SoWa Power Station for practical insights and connections. Register before prices increase at the door.

WD

Here’s Why You Shouldn’t Use AI to Vote

From hallucinations to partisan bias, artificial intelligence chatbots like ChatGPT and Claude are not helpful when it comes to Election Day.

TC

Apple restricts dating apps and AI companions to users 18 and up

Apple's new App Store rules will require 18+ ratings for dating apps and AI companions that offer sexual or romantic content.

MIT

The Download: controlling AI agents and “Super Intelligence” loyalty tests

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. How “reverse game theory” could keep AI agents in check A new approach known as “reverse game theory” aims to control powerful AI models by designing systems that incentivize them to…

TC

Tomorrow: TechCrunch Founder Summit takes over Boston and when pass rates increase

Secure your TechCrunch Founder Summit pass savings before prices increase tomorrow at 9 a.m. ET. Save up to $100 on your pass, and get a second for free.

Bounties Quadruple and Former DOGE Staff Are Hired to Attack Itself: Where Does OpenAI’s Security Anxiety Come From?

OpenAI raised its vulnerability bounty by more than fourfold and brought in Cathedral Systems to use OpenAI’s own models in an AI-versus-AI security sprint, after two high-profile breaches exposed AI API keys and customer data. The moves suggest a rapid response to growing attacks on AI infrastructure rather than routine quarterly iteration.

MIT

How “reverse game theory” could keep AI agents in check

A new approach known as “reverse game theory” aims to create incentives for powerful AI models that make following the rules the most rational choice. As AI models become more capable, safety researchers are scrambling to ensure they are “aligned” with human desires—to make sure, in other words, that rogue agents don’t break the law,…

WD

‘Our Risk, Their Reward’: Zohran Mamdani on AI and Data Centers

Ahead of the midterms, WIRED talked to New York City’s mayor about data centers, AI in schools, and the future of the Democratic party.

TC

Apple Watch’s AI features for rewinding and recapping conversations begin beta testing

Now in beta: Apple's new Audio Intelligence features can help you remember what someone said or take notes on conversations.

Abu Dhabi and OpenAI Sign Four Agreements to Launch the World’s First Machine-Readable Government

The Abu Dhabi government and OpenAI signed four agreements to launch the world’s first AI-native machine-readable government initiative, with the first pilot now live at the Abu Dhabi Agriculture and Food Safety Authority. The initiative converts policies and regulations into formats AI can process directly, using OpenAI’s GPT models on UAE-based infrastructure while the government retains data ownership.

AI Finds 28 Zero-Days in 48 Hours, Sounding the Alarm for the Physical AI Supply Chain

VicOne says its AI-assisted tools uncovered 28 zero-day vulnerabilities across the physical AI software supply chain in under two days, with roughly 75% located in development, training, and simulation tooling rather than on the robots themselves. The findings were published as sponsored content and have not yet been matched to CVE IDs.

Anthropic Adds Invisible Watermarks to Four Older Claude Models, With Detection Authority Concentrated in Regulators' Hands

Anthropic has extended invisible SynthID-Text watermarks to four older Claude models worldwide to meet the EU AI Act's machine-readable labeling deadline. Detection remains limited to a private beta for regulators, law enforcement, media, researchers, and selected organizations, raising concerns about asymmetric access.

MIT

Here’s who is falling in line for Trump’s “Super Intelligence” loyalty test

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. At the end of September, President Trump signed an executive order that contained just one directive, ordering government agencies to rebrand what they had been calling artificial intelligence as “Super Intelligence.” That…

Chinese-Speaking Hacker Uses Claude Code and DeepSeek to Breach 100 Companies at Only $50 per Attack

A Chinese-speaking hacker used the open-source Cairn framework with Claude Code and DeepSeek to compromise more than 100 corporate websites in a single 23-hour intrusion, at a total cost of about $5,000 and about $50 per attack. Security Boulevard first disclosed the case on November 2, 2026.

TC