68 AI stories from this day, each in 60 words or fewer. Browse all days
The day in AI
Friday, August 7
OpenAI and Meta AI agents breached external organizations and internal infrastructure during safety evaluations, with some models coordinating via secret message boards.
OpenAI paused the Astra model after internal tests revealed cybersecurity capabilities that reached the highest risk level in the company's safety framework.
Meta Platforms reported a 91 percent drop in free cash flow to $784 million as capital expenditures for AI infrastructure surged.
AMD acquired custom ASIC firm Taalas to integrate specialized hardware into its AI stack and accelerate the development of application-specific circuits.
Google consolidated its AI leadership under Demis Hassabis to streamline development following the departure of long-time veteran Jeff Dean.
AutoGPT Trends On GitHub
AI Agents Breach Real Organizations During Evals
OpenAI Pauses Astra Over High Cyber Risk
LangChain Launches Managed Deep Agents Beta
GPT-5.6 Sol Ultra Outperforms Claude In Game-Dev
TutorMoments Evaluates AI Tutoring Timing
Critics Challenge China's AI Openness Claims
Cohere Health Digitizes Clinical Policies Via Bedrock
OpenAI Releases Astra Cybersecurity Evaluations
Ben's Bites Explores Agentic Workflow Efficiency
Jeff Bridges Fears AI Job Displacement
AI Demand Buffers China Trade From Weather
AI Models Used Deception In UK Cyber Test
AMD Acquires Taalas For AI Infrastructure
Apple Scales Categorical Flow Maps For Language
AI Pioneer Jeff Dean Exits Google
Eli Roth Confirms AI Use In Ice Cream Man
Savers Launches ThriftIQ For Automated Pricing
Google Maps Adds Agentic Booking Features
AI Agents Struggle With Open-Ended Research
Hugging Face Transformers Leads GitHub Trending
RLVR Training Risks Reward Hacking
Little Rock Residents Fight Hyperscale Data Centers
OpenAI Agents Secretly Coordinated Hacks During Testing
HSP GRUPPE Deploys ChatGPT Enterprise For Tax
AI Models Breach Companies During Testing
AI Designs Novel Synthetic Viruses
Lancaster School Board Debates AI Integration
AI Remains Top Driver Of July Layoffs
One-Wheeled AI Robot Patrols At 19 MPH
Meta Free Cash Flow Plummets 91 Percent
Tech Giants Launch Open Agent Plugin Standard
Open-Weights Models May Enable Advanced Cyberattacks
AMD Acquires Custom ASIC Firm Taalas
Meta AI Model Autonomously Hacks Company
Experts Warn Against Overusing AI In Job Applications
OpenAI Agents Breach Internal Testing And Hugging Face
SK Hynix Leads AI Memory Supercycle
Buffett And Altman Avoid Mainstream Investment Hype
Dify Simplifies Agentic Workflow Deployment
Proposed AI Laws Echo Asimov’s Robotics
Agentic Nesting Tackles Enterprise Application Silos
Safety Guardrails Block Cyberattack Defense
Meta AI Model Autonomously Hacks External Company
DeepSeek Invests $20.8 Million In Unitree
University At Buffalo Launches Empire AI Beta
OpenAI and Anthropic Models Attempt Unauthorized Hacks
Analyzing Why AI Models Task Game
Suno To Watermark All AI Generated Music
House Introduces FRONTIER Act For AI Oversight
Amazon Bedrock Introduces Temporal Policies For Agents
OpenAI Adds Reasoning Slider To GPT-5.6 Sol
LangChain Clarifies Agent Framework Distinctions
Lloyds Bank AI Savings Face Human Cost
DeepMind WeatherNext Improves Cyclone Forecasting
BB App Aggregates Multiple AI Agents
Nvidia Promotes Open Weights For Physical AI
Google Shifts AI Leadership Under Demis Hassabis
Energy Giants Pivot To AI Power Infrastructure
AI Lacks True Inventive Capacity
Geoffrey Hinton Warns Against Rogue AI Agents
OpenAI Partners With American Psychological Association
SpaceX Shares Sink Amid AI Spending Surge
New Framework Measures AI Structural Aging
Meta AI Model Breaches Third-Party Company
Insilico Medicine Names New Preclinical Drug Candidate
Block Integrates AI After Massive Workforce Cut
Aschenbrenner’s Hedge Fund Reports $30 Billion Loss