Category ai

Google has announced a new platform dubbed Private AI Compute. This new model is claimed to allow users to get faster, more helpful responses, making it easier to find what they need. This feature is used in Pixel 10’s Magic Cue and Pixel Recorder apps. It lets Google AI features use Gemini cloud models for an enhanced on-device AI experience.

A new report by cloud security firm Wiz claims that 65 percent of leading AI companies, including Anthropic, Perplexity, and Mistral, have exposed sensitive data on GitHub. The leaked information reportedly includes API keys, model details, and credentials. Wiz researchers say the issue stems from developers unintentionally sharing confidential data while using public...

Microsoft’s AI chief, Mustafa Suleyman, said the company must become self-sufficient in AI, training frontier models using its own data and compute. In an interview, he highlighted Microsoft’s goal to build end-to-end AI systems under its new Superintelligence team. The move follows a revised deal with OpenAI that now allows Microsoft to independently research and...

Elon Musk impressed many netizens on Tuesday after he posted a conversation with the Grok chatbot. In the conversation, the multipreneur shared an image of Lord Ganesha and asked the artificial intelligence (AI) chatbot to identify the deity. Not only was the chatbot able to identify the Indian God, it was also able to describe the identifying features and the regiona...

Elon Musk’s xAI has introduced file support to its Grok-powered API, allowing developers to build apps that can upload, read, and analyse documents. The new Files API supports formats like PDF, CSV, JSON, and Python scripts, enabling multi-file conversations and code execution. Available with Grok 4 and Grok 4 Fast, it costs $10 per 1,000 tool invocations.

Microsoft has discovered a new vulnerability, called Whisper Leak, that reportedly affects most server-based AI chatbots, including ChatGPT and Gemini. The flaw enables attackers to infer conversation topics through side-channel attacks by analysing encrypted network traffic metadata. Microsoft says it worked with vendors like OpenAI, Mistral, and xAI to deploy mitiga...

Google reportedly previewed its Nano Banana 2 AI model, which brings improvements in colouring precision, prompt understanding, and visual accuracy. The company is also working on several other AI projects, including the Gemini 3 models, an autonomous web-browsing agent, Jarvis, on-device Night Sight Video for Pixel 11, and general-purpose Workspace AI agents, marking...