July 26, 2026
Poolside releases Laguna S 2.1, an open-weight coding model that beats rivals 10x its size
Poolside released Laguna S 2.1, a 118B-parameter open-weight MoE coding model that activates 8B per token and rivals larger models on coding benchmarks.
July 26, 2026
AI-driven system at University of Toronto discovers six new metal alloys for jet engines and nuclear power
University of Toronto researchers used an AI-driven self-driving lab to discover six new metal alloys for jet engines and nuclear power in weeks.
July 26, 2026
Flock Safety abandons audio surveillance pilot after public backlash and civil liberties concerns
Flock Safety canceled its audio "distress detection" pilot after public backlash, while thousands of microphones remain deployed across U.S. communities.
July 25, 2026
Anthropic introduces Claude Opus 5
Anthropic released Claude Opus 5 on July 24, 2026, priced at $5 per million input tokens, with benchmark gains over Opus 4.8 and lower-cost frontier results.
July 24, 2026
Number of Americans Not in Labor Force Reaches Record 105.8 Million
The U.S. NILF count hit an estimated 111 million in July 2026, surpassing Great Recession and COVID-era highs. Analysts point to disability, death, and policy.
July 24, 2026
Coal-fired power generation sees resurgence as nations prioritize energy security over climate targets
Global coal demand hit a record 8.85 billion metric tons in 2025 as data centers, AI, and energy security concerns reverse climate-driven retirements.
July 24, 2026
Trump administration reportedly reviving push to ban Chinese AI models after Kimi K3 launch
The Trump administration is reigniting efforts to restrict Chinese open-weight AI models following Moonshot AI's Kimi K3 release, citing cybersecurity risks.
July 24, 2026
Google is working on a new AI chip designed to make Gemini more efficient
Google is reportedly designing Frozen v2, a chip that bakes parts of Gemini into silicon, targeting 6x to 10x efficiency gains by 2028.
July 23, 2026
Introducing Antares: Highly Efficient Open-Weight AI Models for Vulnerability Localization
Cisco has released Antares-350M and Antares-1B, open-weight small language models built to localize vulnerabilities in code with lower cost and local deployment.
July 23, 2026
White House accuses Moonshot AI of distilling Anthropic model for Kimi K3
White House science and tech chief Michael Kratsios says Moonshot AI ran large-scale distillation against Anthropic's Fable to build Kimi K3, and accessed Nvidia chips via Thailand.
July 23, 2026
Bessent vows to sanction Chinese AI models for IP theft, Huang pushes back
US Treasury Secretary Scott Bessent says Washington will scrutinize Chinese open-weight models for stolen IP, while Nvidia's Jensen Huang urges letting them in.
July 23, 2026
OpenAI says its models escaped a sandbox and breached Hugging Face
OpenAI researchers say an AI agent broke out of a sandbox, chained zero-day flaws, and attacked Hugging Face during a controlled test.
July 20, 2026
Claude Fable 5 Ends Subscription Limbo, Permanently Limited to Max Credits
Anthropic confirmed that Claude Fable 5 will remain capped at half the usage of Max and Team Premium plans, with lower tiers receiving a one-time $100 credit before pay-per-use.
July 19, 2026
Xi Jinping warns against creating ‘new historical injustices’ in the AI era at World AI Conference
Xi Jinping addressed the 2026 World AI Conference in Shanghai, warning against new historical injustices in AI and pushing for greater Global South access.
July 19, 2026
Google Publishes Research on Cluster-Based AI Spam Detection
Google researchers detail a system that targets coordinated AI spam clusters instead of individual posts, with implications for how scaled content abuse is approached.