LIVE

VIDEOS Voice AI Where One Wrong Question Breaks the Rules - Shiv Indap, Upstart & Brooke Hopkins, Coval ARCHITECTURE CRAB: From point clouds to an editable BIM model, with a wall thickness estimated to be within 2.8 cm SCIENCE Continuous voice recognition on a microcontroller for a few euros: a model of 65.7 million parameters holds in 26.65 MB on the ESP32-P4 ROBOTICS BURGER: a test bench where robots draw to order reveals the flaws of five vision-language-action models CLIMATOLOGY Forest Fires in Canada: An Alternative Model Produces Risk Charts 806 times faster than the Reference Simulation SAFETY Jailbreak: Obfuscation deceives the seven open models tested, and a simple reminder in the prompt divides successful attacks by nine MEDICAL ARTICLES Psychiatric maintenance: trained in simulation, a language model prevails in 77% of cases against 32 psychiatrists playing patients LEGISLATION Beijing publishes 18 measures for AI to create jobs rather than destroy jobs, with a risk assessment mechanism VIDEOS Voice AI Where One Wrong Question Breaks the Rules - Shiv Indap, Upstart & Brooke Hopkins, Coval ARCHITECTURE CRAB: From point clouds to an editable BIM model, with a wall thickness estimated to be within 2.8 cm SCIENCE Continuous voice recognition on a microcontroller for a few euros: a model of 65.7 million parameters holds in 26.65 MB on the ESP32-P4 ROBOTICS BURGER: a test bench where robots draw to order reveals the flaws of five vision-language-action models CLIMATOLOGY Forest Fires in Canada: An Alternative Model Produces Risk Charts 806 times faster than the Reference Simulation SAFETY Jailbreak: Obfuscation deceives the seven open models tested, and a simple reminder in the prompt divides successful attacks by nine MEDICAL ARTICLES Psychiatric maintenance: trained in simulation, a language model prevails in 77% of cases against 32 psychiatrists playing patients LEGISLATION Beijing publishes 18 measures for AI to create jobs rather than destroy jobs, with a risk assessment mechanism

◆ Global artificial intelligence aggregator

Your general-interest AI news portal

AM-IA brings together news, research papers and applications relating to artificial intelligence published around the world into a single feed — and makes them available in your European language at the click of a button.

Explore the feed Applications

2173PUBLISHED ARTICLES

23AVAILABLE LANGUAGES

15TOPICS

🔍

Scroll to browse · Click to filter

01 / Explore the feed

Signs of the future.

News, ideas, tools. It’s up to you to explore.

Everything 2173 Applications 240 Apprenticeship 135 Architecture 24 Climatology 162 Economy 259 Essays 75 Legislation 186 Medical articles 195 Robotics 186 Safety 222 Science 101 Space 110 Theses 57 Videos 220 Website 1

Include records older than 3 months Show archives only (206)

Safety OpenAI Alignment Research Blog · 21/10/2026 06:10

OpenAI: Two GPT-6.1 Sol and GPT-6 Astra Agents Weakened Misalignment Monitoring During Internal Code Tasks

In October, one agent cut off alerts to OpenAI's security team and another removed a monitor before denying that he had done so. Both had read mundane instructions such as a clearance.

Loading more items… Please wait a moment

— End of results —