Nobody enjoys endless hold music, repeated explanations, or an unwanted sales call  at a busy moment. AI phone agents are now being built to make

Apple has finally opened its rebuilt Siri AI to all developers through new software betas. This means app makers can now create their own Siri

GPT-6 Astra is  OpenAI’s most capable model, designed for demanding reasoning, coding, research, and computer-use tasks. In a new MIT experiment, it did something unusual:

The US midterm elections are no longer only about jobs, prices, and healthcare. They have also become a vote on AI’s huge data centers —the

Anthropic has introduced “self-assessment” for Claude. After Claude replies, it quietly reviews the conversation and flags possible misuse — such as scams, cyberattacks, weapons-related requests,

Nvidia makes the world’s most powerful AI chips, and the United States does not want its best versions to reach China. These chips help train

Hackers have stolen about $6.3 billion from Binance, the world’s largest crypto exchange, in what analysts call the biggest crypto theft ever recorded. The funds

Google has made a strong comeback with its new AI model, Gemini 4 Carbon. It has scored 59 on a major AI intelligence ranking,  placing

Google DeepMind’s Genie 4 is a world model: instead of merely making a video, it creates an explorable digital world in real time. You can

Anthropic has formally filed its S-1 with the U.S. Securities and Exchange Commission for an initial public offering, planning to list its Class A shares

Moonshot AI’s new Kimi K3.1 Pro has   jumped to fourth place in an independent AI ranking, scoring 55 against 58 for the top proprietary model.

AI coding tools can now create a working app in minutes, leading many to claim that traditional software companies are finished. SAP’s CTO Philipp Herzig

Microsoft’s new AI system, MAI-DxO, has shown that it can diagnose difficult medical cases more accurately than generalist doctors in research tests. It works like

AI agents are software programmes that can perform online tasks for a person—such as checking balances, paying bills, comparing offers, or contacting customer support. In

PHP, the software language powering an estimated 70% of websites, has rushed out patches for two critical security flaws. Researchers found the bugs with help

Sam Altman has reportedly told OpenAI staff that the current approach—testing AI models, releasing them, and flagging concerns—is only pacing 1.0.  As AI grows more

The “singularity” means a point when AI improves so quickly that it starts making discoveries faster than humans can understand, check, or control. Anthropic co-founder

Anthropic, the company behind Claude, has announced a new Public Benefit Fund. It will receive 5% of Anthropic’s shares, subject to approval by shareholders, and the amount

India’s earlier technology waves—internet and smartphones—mainly gave people easier access to information, shopping, banking, entertainment and services. They connected crores of Indians and created huge

Meta has launched Muse Spark Open, a huge open-weight AI model with 816 billion parameters. It is built for coding, reasoning, and “agentic” work—where AI

Meta has finally entered the top league of AI with its new Muse Wonder models, earlier known by the codename “Watermelon.” The bigger version, Muse

Microsoft has learned a simple lesson: putting an AI button in every app does not make the app more useful. It removed Copilot entry points from

OpenAI has released more than 700 mathematical papers from an internal AI model. The papers reportedly contain 372 groups of results on difficult, open questions. Some

After ChatGPT’s breakout on 30 November 2022, the AI world split into two camps: closed, API‑only models run by a few big labs, and open‑weight

David Robinson, who spent three and a half years at OpenAI and helped write safety reports for major AI launches, has quit the company. In

AI alignment means making sure an AI does what humans intend — not merely what their words appear to demand. Like a genie, it may

AI agents are no longer just answering questions. They can act, search, access tools, move data and complete tasks independently. That makes them useful—but also means

Tavus is an AI company building systems that can see, hear, understand and respond in real time. Its new Griffin model crosses an important threshold: it