Category: AI News
-
A recent blockchain AI bug exposed a major weakness in smart contract technology. The flaw created multi-million-dollar risks for Web3 platforms and raised serious concerns about the growing reliance on artificial intelligence in blockchain security. The Bug Discovery Cybersecurity researchers uncovered the blockchain AI bug in a widely used protocol. The flaw allowed attackers to…
-
The Perplexity Comet browser has officially become free for all users. The AI-powered browser, once limited to a $200 monthly plan, now offers open access with usage limits. The move aims to expand Perplexity’s user base and challenge Chrome’s dominance in the AI browsing space. AI-Powered Features Built on Chromium, the Perplexity Comet browser combines…
-
A new study shows that AI writes UN press releases more often than many realize. Researchers found that around 13% of official UN statements now contain markers of AI involvement. The discovery highlights the rapid spread of artificial intelligence into global institutions and raises questions about transparency in public communications. Findings From the Study The…
-
Google has taken action against Trump-related AI search features, blocking access after citing repeated policy breaches. The decision highlights the growing tension between political figures and technology platforms over how content is moderated and represented in artificial intelligence. Why Google Blocked Access According to Google, the Trump-related AI search results failed to comply with its…
-
The Wikidata embedding project aims to transform Wikipedia’s structured data into a powerful tool for artificial intelligence. Wikimedia Deutschland has begun converting 120 million open data points into vector embeddings, making this vast knowledge base more accessible to AI systems. How the Project Works The project turns Wikidata’s statements into vector representations. These embeddings are…
-
Meta is preparing to introduce Meta AI targeted ads, a new advertising approach that uses AI chat data for personalization. The move highlights Meta’s continued reliance on user information to fuel its business model, but also raises concerns over privacy and consent. How Meta AI Will Target Users Starting December 16, 2025, Meta will collect…
-
Researchers have discovered serious Gemini security flaws that highlight how attackers can exploit Google’s AI system. These weaknesses, known as the “Gemini Trifecta,” reveal gaps in prompt handling, data protection, and personalization features. The Three Major Vulnerabilities Experts identified three separate flaws in Gemini. Each posed a risk of data theft or manipulation: How Prompt…
-
The Amazon AI-powered Alexa+ has been unveiled alongside a refreshed lineup of devices. At its latest hardware event, the company introduced updated Echo speakers, Fire TVs, Ring and Blink cameras, and a new Kindle. With this update, Amazon aims to make Alexa+ more conversational, responsive, and central to smart home life. Smarter Echo Devices The…
-
The DeepSeek experimental AI model has been unveiled in China with bold promises. It aims to deliver faster performance while cutting costs by half. With AI tools becoming increasingly expensive to run, this launch signals a shift toward efficiency-driven innovation. What the Model Offers DeepSeek introduced DeepSeek-V3.2-Exp on Hugging Face as an upgrade to its…
-
The California AI regulation bill has been signed into law by Governor Gavin Newsom. The legislation introduces accountability rules for AI developers, including industry leaders like OpenAI and Google. As a result, California is positioning itself as a frontrunner in setting standards for artificial intelligence oversight. New Rules for AI Giants Under the new law,…