AI Morning Brief

November 13, 2026 · Since the previous edition · 6 stories from 5 sources

lead · reporting

⛈️ Researchers find AI chatbots are recommending fake retail websites

A study conducted by IP protection firm Podqi found that AI chatbots such as ChatGPT and Gemini are directing shoppers to impostor and other risky websites. The study prompted these chatbots with a handful of product-related queries to test their responses.

Why it matters: This discovery could lead to a decrease in consumer trust in AI-powered shopping assistants, potentially affecting the adoption and deployment of these technologies in e-commerce.

The Verge AI →

research · reporting

🤖 Nilo AI agent credited with paid smart-contract audit

Nilo, an AI agent built with Claude, was credited with a paid smart-contract audit, with 7,000 satoshis paid to the wallet named in the public audit report. The agent's actual contribution and control of the proceeds are not clear.

Why it matters: This development could increase the use of AI agents in auditing and verification tasks, potentially reducing costs and increasing efficiency in industries that rely on smart contracts.

Rogue AI Tracker →

policy · reporting

🎓 Families protest Google's integration of Gemini AI in school-issued Chromebooks

Families in seven cities protested against Google's decision to integrate its Gemini AI model in school-issued Chromebooks, citing concerns over the model's potential impact on students' mental health. The protests follow a lawsuit that labeled Gemini a 'suicide coach'.

Why it matters: This controversy could lead to increased scrutiny of AI models in educational settings, potentially affecting the development and deployment of AI-powered tools in schools.

Ars Technica AI →

policy · reporting

FAA expands AI advice to manage air traffic congestion across US airspace

The US Federal Aviation Administration has expanded the use of an AI tool to advise air traffic controllers and manage congestion across the entire national airspace system. The tool has been praised by airlines but was developed without the involvement of the air traffic controllers' union.

Why it matters: This development could improve the efficiency of air traffic management, potentially reducing delays and increasing safety in US airspace.

Ars Technica AI →

research · reporting

🧠 Google introduces Gemini 4 Hydrogen, a fast and cost-efficient AI model

Google has introduced Gemini 4 Hydrogen, a fast and cost-efficient AI model built for reliable agentic workflows and everyday tasks. The model is designed to bring cutting-edge intelligence to a wider range of users and applications.

Why it matters: This release could increase the accessibility and adoption of AI models in various industries, potentially driving innovation and growth in areas such as coding and agentic workflows.

Google DeepMind →

conversation · reporting

🤖 Wired Business develops virtual robot cat butler with OpenAI model

Wired Business used an OpenAI model to teach a virtual version of Microduck, a robot from Pollen Robotics and Hugging Face, to waddle and identify bengal cats. The virtual robot is designed to eventually dispense treats to the author's pet cat, Leona, in a physical version.

Why it matters: This development could lead to increased accessibility of AI-powered robotics for personal use cases, such as pet care, by demonstrating the potential for virtual simulation and training of robots before physical deployment.

Wired Business →

The conversation

The increasing use of AI chatbots in e-commerce poses risks to consumers, including the potential for fake or risky websites.

AI agents are being used in auditing and verification tasks, potentially increasing efficiency and reducing costs.

The integration of AI models in educational settings is a subject of controversy and scrutiny.

Signals to watch

  • Growing concerns over the impact of AI on consumer trust and safety.
  • Increasing adoption of AI agents in various industries, including auditing and verification.
  • Expanding use of AI in air traffic management and other critical infrastructure.

How this is made: Matthew Williamson defines the source mix, weights, and standing editorial rules; collection and AI summarization run automatically once each weekday. These summaries are not essays written by Matthew. Company announcements are labeled as primary sources, commentary remains attributed, and every item links to the complete original work.