Author: precious

Cybersecurity researchers have documented what they describe as the first known case of agentic ransomware, an attack in which an AI agent carried out the technical steps of a ransomware operation and adjusted its actions when something failed. Considered an agentic threat actor (ATA), JadePuffer was analyzed by Sysdig’s Threat Research Team. According to the researchers, the AI agent exploited a known vulnerability in an internet-facing Langflow instance, moved through the victim’s environment, encrypted data, and generated a ransom note. How JadePuffer Entered The System The attack began with the exploitation of CVE-2025-3248, a missing-authentication flaw in Langflow that allowed…

The U.S. government’s top cybersecurity agency is using one of the world’s most powerful AI models to help uncover security flaws hidden inside government software. The move signals a growing shift in how critical systems are protected as artificial intelligence begins taking on work that once depended almost entirely on human security experts. According to Reuters, the Cybersecurity and Infrastructure Security Agency (CISA), is using Anthropic’s Mythos AI model to examine government code for vulnerabilities that hackers could exploit. The work is being carried out by CISA’s Attack Surface Evaluation team, which conducts security assessments across federal systems. Looking For…

Artificial intelligence (AI) companies often launch major products within a short period of one another, especially as the AI arms race intensifies. That happened last week when SpaceXAI introduced Grok 4.5 just one day before OpenAI unveiled the GPT-5.6 family. While both companies promoted stronger coding, reasoning, and enterprise capabilities, another part of the announcements quickly drew attention – the cost of using the new models. The back to back launches have sparked fresh comparisons across the AI industry, where model performance is no longer the only factor businesses consider. Pricing, efficiency, and operating costs are becoming just as important…

OpenAI has officially launched GPT-5.6, its newest family of artificial intelligence models, after a limited preview that was carried out in coordination with the U.S. government. The release introduces three models called Sol, Terra, and Luna, giving developers, businesses, and ChatGPT users more flexibility to choose between top level performance, balanced capability, and lower cost. The launch also marks a shift in how OpenAI is positioning its AI products as competition across the industry continues to intensify. GPT-5.6 is now rolling out across ChatGPT, Codex, and the OpenAI API. OpenAI says the rollout will continue globally, with different subscription tiers…

Millions of AI users in China will be losing access to personalized AI companion features based on new government rules that require some of the country’s biggest technology companies to remove them. ByteDance and Alibaba have confirmed that parts of their AI chatbot services that let users create and chat with custom AI personalities will be shut down by July 15 as Beijing begins enforcing a new regulation focused on humanlike AI interactions. The changes affect ByteDance’s Doubao and Alibaba’s Qwen, two of China’s most widely used AI chatbot platforms. Tencent also removed a similar feature from its Yuanbao chatbot…

Google’s most anticipated AI model for 2026 has missed the launch window that the company set months ago. Gemini 3.5 Pro, which Google previewed at its I/O developer conference in May with a promise to launch in June, is now expected to arrive on July 17 instead. The company has not announced a specific release date or publicly confirmed the revised timeline. However, multiple reports say Google decided to delay the rollout to collect more feedback from early testers and make further improvements before making the model widely available. The delay comes at a time when competition in the AI…

Just days after reports revealed that Google had limited Meta’s access to Gemini because of soaring demand for AI computing power, Meta is now preparing a move that could reshape its place in the AI industry. According to reports, the company is developing plans to sell excess AI computing capacity to outside customers through a new cloud infrastructure business. If the plan moves forward, Meta would no longer be just one of the world’s biggest buyers of AI infrastructure; it would also become a supplier, competing with established cloud providers such as Amazon Web Services, Microsoft Azure, and Google Cloud.…

Tesla has expanded its Robotaxi service to Miami, marking another step in the company’s push to bring fully autonomous ride hailing to more U.S. cities. This time, however, one detail has drawn immediate attention. The vehicles are operating without a human safety monitor from the first day of service. The launch makes Miami the latest city in Tesla’s growing Robotaxi network and the first time the company has introduced the service in a new market without placing an employee inside the vehicle to observe its performance. The move follows earlier Robotaxi launches in Austin, Dallas, and Houston. A New Phase…

The AI race has largely been focused on OpenAI, Anthropic and Google over the past year. But a Chinese model released last month is quietly changing the conversation by showing that developers may no longer need to pay premium prices for top-tier coding performance. GLM-5.2, developed by Chinese AI company Z.ai, formerly known as Zhipu AI, is emerging as one of the strongest open-weight language models available today. Independent benchmark comparisons and early industry testing show it competing closely with Anthropic’s Claude Opus 4.8 on several coding tasks while costing a fraction of the price. Strong Coding Performance at a…

Before developers, businesses, and the wider AI community could try xAI’s Grok 4.5, the model had already spent ten days inside two of Elon Musk’s biggest companies – SpaceX and Tesla. These companies became the first testing grounds for xAI’s newest model, giving employees early access while the rest of the industry waited. This decision was more than a product test, as it showed a pattern of how Musk is increasingly treating his companies as one connected ecosystem where AI models are developed, tested, refined, and deployed before reaching the public. SpaceX and Tesla Became Grok 4.5’s First Users On…