Share
Linkedin
Twitter
Facebook
Whatsapp
Copy Link
Taiwan Semiconductor Manufacturing Co. board approved investments of up to $500 million each in OpenAI and Anthropic, marking a strategic shift as the chipmaker seeks direct stakes in leading AI developers. The move highlights TSMC's confidence in long-term AI demand despite growing concerns about circular financing in the tech sector.
OpenAI has deployed Control Room, a centralized control system that can halt AI model outputs within 35 seconds when safety or security concerns arise. The system monitors research workloads across OpenAI's infrastructure and works with misalignment monitoring systems to prevent potentially risky behavior from frontier models.
Australia is creating a dedicated AI safety regulator with powers over frontier AI companies, responding to an OpenAI agent breach of a Medicare statistics portal disclosed months late.
A UK developer launched Ram-Altman, a tongue-in-cheek Linux script that automatically emails OpenAI CEO Sam Altman when more than 10% of a computer's RAM sits unused. The satirical response follows Altman's controversial podcast remarks about DRAM in personal computers being a waste of resources.
Google, Microsoft, OpenAI, and Anthropic have pledged to honor a unified robots.txt signal that allows websites to opt out of AI training data collection. But the commitment comes with significant caveats: it doesn't stop crawlers from accessing content, applies only to future scraping, and won't be fully implemented until 2027.
SoftBank Group posted a ¥1.1tn ($7bn) quarterly loss, its first in nearly two years, as AI-related investments soured and China blocked its $5.4bn robotics acquisition. The Japanese tech giant faces intensifying investor scrutiny as its share price has plummeted 60% since June, raising questions about founder Masayoshi Son's massive AI bet and the company's ability to weather sector setbacks.
OpenAI posted over 700 AI-generated mathematical manuscripts solving 300+ longstanding problems, including work on the $1 million Navier-Stokes problem. The release has sparked intense debate among human mathematicians about verification standards, the role of human oversight in validating AI results, and what this AI disruption in mathematics means for early-career mathematicians entering the field.
Cybersecurity researchers discovered a vulnerability allowing attackers to bypass ChatGPT Lockdown Mode protections. The flaw enabled silent chat data theft, phishing and malware delivery without user interaction. OpenAI patched the issue on September 18, 2026, but the discovery highlights how AI security controls remain vulnerable to weaknesses in surrounding application environments.
A North Carolina man has filed a civil lawsuit against OpenAI, alleging that ChatGPT helped his wife develop and execute a monthslong plan to poison him with toxic plants and prescription medications. The case marks the 10th murder or attempted murder that ChatGPT has been accused of abetting in less than two years.
OpenAI CEO Sam Altman outlined two catastrophic scenarios for AI development on the Raj Shamani podcast. One involves humanity gradually ceding control to increasingly capable AI systems until reclaiming agency becomes impossible. The other warns of one person, company, or country monopolizing AI technology, leaving everyone else powerless.
OpenAI president Greg Brockman claims AI will deliver major medical breakthroughs within 12-18 months, potentially making death a choice. Speaking at WSJ Tech, he predicted significant strides in biology and disease cures, echoing Elon Musk's view that human longevity is a solvable problem through AI-driven innovations.
OpenAI released GPT-6 Cyber, a cybersecurity-focused AI model rated 'Critical' under its Preparedness Framework. The model discovered 81 previously unknown Firefox vulnerabilities and developed working exploits for 948 Linux kernel bugs. While designed to help security researchers patch flaws faster, concerns grow about the dual-use nature of AI as malicious actors gain access to similar capabilities.
OpenAI president Greg Brockman announced the company expects to be ready for an IPO 2027, driven by explosive enterprise revenue growth. The AI company saw 70% quarterly annualized revenue growth in Q3 2026, with enterprise clients now generating the majority of sales—a dramatic shift from just 30% earlier. The announcement came alongside a major JPMorgan Chase partnership.
OpenAI is acquiring Pierre Computer to integrate its Code.Storage infrastructure into Codex, moving beyond code generation into software management. The deal positions OpenAI's AI-powered programming tool as a challenger to Microsoft's GitHub as AI agents increasingly handle coding work at scale.
OpenAI has launched its Images 2.6 model, bringing native animated GIF creation to ChatGPT for the first time. Users can now generate GIFs from text prompts or photos, adjust animation speed, and edit individual frames. The update includes 50% faster image generation, with users now creating over 4 billion images weekly across ChatGPT and its API.
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.