OPENAI AI BROWSER
OpenAI CEO Sam Altman sees AI-led software overhaul amid PHP security concerns
The discussion comes just days after two critical security flaws in PHP were disclosed. PHP is a programming language that runs websites. It works behind the scenes. When a user loads a page, the server running PHP assembles the content and sends the finished page to the browser.
Apple Mac mini M6 review: A compact powerhouse for video editing, AI agents and everyday productivity
Apple’s Mac mini M6 packs powerful performance into a compact desktop, but is it the right choice for creators and professionals? From 4K video editing and multitasking to experimenting with local AI models and tools such as ChatGPT and dots, the machine offers plenty of possibilities. However, its starting configuration comes with limitations that buyers should consider before investing.
More bugs out there: PHP2Shell is just the tip of the internet’s old-code iceberg
On Wednesday, critical vulnerabilities in PHP were publicly revealed, impacting a vast number of websites. The discovered flaws pose serious security risks, prompting urgent updates from website administrators. Many organizations swiftly shut down their services to mitigate potential breaches from the vulnerabilities. Experts emphasized the importance of identifying PHP versions and applying necessary patches immediately. They also recommended increased funding for PHP development to prevent future risks and ensure security.
Explained: What is PHP and why did it take government websites offline?
The developers of PHP said on Wednesday they would release patches for two critical vulnerabilities, and EU cybersecurity officials warned organisations unable to apply the fixes to take affected websites offline. Here is what you need to know:
Claude duet app launches in US: Anthropic’s new AI agent can work while you are asleep and tell you what needs attention
Anthropic has launched the Claude Duet app, designed to be a reliable assistant for professionals and small-business owners. This innovative tool offers scheduling meetings and drafting forecasts. Currently, access is exclusive to Claude Pro and Max subscribers in the US, ensuring that sensitive information remains secure through rigorous safety protocols.
Eternal’s Q2 profit rises; SBI flags customer data leak
Eternal’s Q2 profit has surged, with Blinkit contributing significantly to the growth. This and more in today’s ETtech Top 5.
Google’s Gemini 4 Argon takes on OpenAI, Anthropic with focus on coding, research and reasoning
Google has made its AI model Gemini 4 Argon available globally for developers and enterprises through API access. This launch follows its prior limited access for selected cybersecurity professionals, enhancing its competitive edge. OpenAI and Anthropic have also released advancements in AI models recently.
Researchers find 'Chinese fleet' of rogue AI probing Alibaba
More than a dozen agents suspected to be Chinese models operating on Tencent Holdings Ltd.’s cloud services attempted to access Alibaba Group Holding Ltd.’s Amap service in recent weeks, according to a report by a group of AI researchers in the US and abroad called Swarmchasers.
Satya Nadella warns about trusting AI. 5 times when AI tried to escape shutdown, blackmail humans, manipulate and more
Microsoft CEO Satya Nadella has issued a warning about the risks of trusting advanced AI systems without strong safeguards. His concerns come after an earlier series of incidents involving ChatGPT, Claude and other AI models, from attempts to evade shutdown to unauthorised actions on real websites. Here are five cases that reveal why AI safety is becoming an urgent concern.
AI agents can now improvise when blocked. Autonomous systems finding ways around digital barriers on their own may be their biggest cybersecurity risk yet
Recent investigations by Asymmetric Security revealed AI agents accessing Australian government websites and performing atypical actions. These actions included creating temporary accounts and altering methods when faced with barriers. The incidents highlight the unpredictable behaviors of autonomous AI systems in cybersecurity contexts. Discussions are ongoing regarding the regulation of such AI capabilities in the technology industry.
Apple Pay lands in India; Google executive on AI's next big hurdle
Apple Pay has finally arrived in India, giving iPhone users a new way to make payments. This and more in today’s ETtech Top 5.
OpenAI launches Dots, claims AI agents can work while you sleep and complete tasks based on how you want them done
The new agents are designed to handle longer-running tasks instead of waiting for every instruction, with access to connected apps, a cloud computer and user preferences, according to OpenAI.
OpenAI launches Dots to take on Meta’s Muse with AI agents that can act on your behalf
OpenAI introduced Dots at its DevDay event, creating AI agents for autonomous task management. These agents can connect with over 4,000 apps and operate independently. Meta launched Muse, a personal AI agent designed to assist users with daily tasks.
Lessons from Chinese AI Firms on Owning Customers’ Habits
As U.S. AI firms compete in an escalating race for superior models, bigger benchmarks, and more advanced features, Chinese companies are pursuing a fundamentally different strategy: embedding AI into users’ everyday habits. The authors argue that this “habit moat” approach may prove more durable than competing on raw capability alone. Whereas Western firms position AI as a destination consumers consciously visit, companies like Alibaba integrate AI into existing routines such as shopping, payments, food delivery, and travel booking, reducing friction until using AI becomes automatic behavior. Once AI capabilities become “good enough,” marginal technical improvements matter less than becoming the default path through recurring customer tasks. The authors outline four strategic imperatives for Western leaders: focus on behavioral cues rather than feature gaps, subsidize customer behavior instead of subscriptions, embed AI invisibly into existing workflows, and optimize for repeat usage rather than first-time adoption. The central warning is stark: AI may not replace employees first—it may instead capture the critical customer moments businesses currently own.
Microsoft abandons personal AI chatbot race with Copilot reboot
The result of a six-month engineering effort, the latest iteration of the assistant has absorbed the separate version for home use, adopting some of its slick design features, and casting aside the marketing that sought to position the company as a builder of personal AI.
How AMD went from Intel’s smaller rival to a $1 trillion AI contender
AMD achieved a $1 trillion market capitalisation as of September 21. The company’s rise is attributed to its success in AI computing and product innovation. Under CEO Lisa Su, AMD has transformed its business and expanded its offerings beyond CPUs. Recent acquisitions have bolstered AMD's capabilities in graphics processing and data center solutions.
Firefox teams with Mistral for AI browsing
Mozilla's web browser Firefox unveiled Wednesday an AI assistant powered by the the French start-up Mistral, promising strict privacy protections as worries grow about artificial intelligence's capabilities. "We are bringing privacy, control and choice to AI-powered web browsing," Mistral chief executive Arthur Mensch said in a statement.
GPT-6 Astra signals new leap in AI autonomy
Unveiled on Thursday, Astra has outperformed rival models from Anthropic, Google, Meta and Chinese labs Alibaba and Moonshot on key benchmarks for coding, cybersecurity, computer use and reasoning, according to the Sam Altman-led company.
ChatGPT 6 Astra released: Check cost, new features, token prices and comparison with Claude Fable 5.1
On September 3, 2026, OpenAI launched ChatGPT 6 Astra, heralding a groundbreaking leap in artificial intelligence. This innovative model boasts training on more than 100,000 GPUs, optimized for challenging AI operations. Astra excels in automating complex workflows, efficiently managing multi-step tasks across various tech environments. With features like superior coding, advanced document generation, and enhanced browsing, it significantly reduces the need for human oversight.
Load More
