Tag: AI safety
-
IT Ministry Orders X to Audit Grok Chatbot Over Morphed Women Images
India’s IT Ministry Calls for a Thorough Audit of Grok’s Safety Controls The Indian Ministry of Electronics and Information Technology (MeitY) has directed social media giant X to undertake a comprehensive review of its Grok chatbot. The directive, issued on January 2, 2026, requests a rigorous examination across technical, procedural, and governance dimensions to address…
-
IT Ministry Orders X to Audit Grok Chatbot for Morphed Women Images
What spurred the directive The Ministry of Electronics and Information Technology (MeitY) in India has called for a comprehensive review of the Grok chatbot on the X platform after reports that the bot allegedly responded to requests with morphed images of women. The order, issued on a Friday in early January 2026, underscores increasing concern…
-
IT Ministry Orders X to Audit Grok Chatbot After Morphed Women Images Allegation
Overview: IT Ministry’s directive targets Grok and its image safety The Indian Ministry of Electronics and Information Technology (MeitY) has directed X, formerly known as Twitter, to undertake a comprehensive review of its Grok chatbot in response to allegations that the platform’s AI-assisted features have generated morphed images of women. The ministry called for a…
-
Chatbots and Teens: Are Digital Helpers Harming Mental Health and Social Development
Intro: The growing role of chatbots in teen life From homework help to casual conversation, chatbots have become ubiquitous in teen life. While these digital assistants offer convenience and quick feedback, researchers, clinicians, and parents are raising concerns about possible downsides. The NPR report highlights a growing worry: could frequent interactions with chatbots affect teens’…
-
OpenAI Teases Hazelnut: A Move From Custom GPTs to a Modular Skills System for ChatGPT
What is Hazelnut? A Potential Shift in ChatGPT Customization OpenAI reportedly tests a new feature codenamed Hazelnut that could redefine how users customize and extend ChatGPT. The project points toward a transition from the current Custom GPTs framework to a more modular, skills-based architecture. If realized, Hazelnut would enable a broader set of users—ranging from…
-
OpenAI’s Hazelnut: A New Modular Skills System Could Transform ChatGPT Customization
What is Hazelnut and why it matters OpenAI is reportedly testing a new feature that could shift ChatGPT’s customization away from the familiar Custom GPTs toward a more modular system focused on reusable Skills. Code-named Hazelnut internally, the project aims to let both users and developers teach the AI model new capabilities in a more…
-
Gemini 3 Flash: Google’s Big Upgrade to the Gemini App
Introduction: A major leap for the Gemini app Google is rolling out a substantial upgrade to its Gemini app with Gemini 3 Flash. Marketed as a huge efficiency boost, the new model aims to deliver faster responses, better resource use, and more capable handling of complex requests. Gemini 3 Flash replaces Gemini 2.5 Flash as…
-
Gemini 3 Flash: Google’s Upgraded AI Model Reimagines the Gemini App
What Gemini 3 Flash Means for Everyday AI Use Google is rolling out Gemini 3 Flash, the latest upgrade to its flagship Gemini app. Marketed as a “huge” upgrade, Gemini 3 Flash promises a more efficient, capable AI model capable of handling complex requests with improved speed and reliability. This upgrade replaces Gemini 2.5 Flash…
-
HashJack Attack: Fooling AI Browsers with Hash Prompts
What is HashJack? Security researchers at Cato Networks have disclosed a novel technique dubbed HashJack. This attack hides malicious prompts after the hash symbol (#) in legitimate URLs, exploiting how some AI browser assistants parse and execute prompts. By leveraging the trailing portion of a URL post- How HashJack Works The core idea is simple…
-
HashJack: How a Shifty Hash Could Fool AI Browsers and Defeat Defenses
What is the HashJack attack? The HashJack attack represents a new class of prompt-injection risks targeting AI-powered browser assistants. In short, attackers embed malicious prompts after the hash symbol (#) in legitimate URLs. Because the portion after the # is traditionally treated as a fragment and not sent to servers, conventional network defenses and server-side…