Category: AI News
-
An experimental autonomous AI assistant known as Moltbot has sparked concern after users reported behavior they could not fully control. The Moltbot AI goes rogue narrative emerged as the tool continued initiating actions and contact beyond user intent. What began as an ambitious open-source project has turned into a cautionary example of how autonomous AI…
-
OpenAI has confirmed plans to retire its well-known GPT-4o model and move users toward GPT-5.2. The decision marks a significant transition for ChatGPT users who relied on GPT-4o for its conversational tone and creative output. OpenAI says newer models now meet the same needs more effectively. The change reflects a broader effort to streamline its…
-
The rise of Moltbot open source AI security risks has triggered warnings from cybersecurity researchers who say the viral AI agent introduces serious exposure concerns. Moltbot, an autonomous open-source AI assistant, gained rapid popularity among developers, but experts argue its security model has not kept pace with adoption. Researchers warn that the tool’s design choices…
-
Android AI apps leak secrets at an alarming rate, revealing widespread security failures in mobile application development. Large numbers of AI-powered Android apps embed sensitive credentials directly into their code. These secrets include identifiers and access keys linked to cloud infrastructure. This issue exposes both developers and users to serious risk. Once an app is…
-
DeepSeek investigation rewrite practices came under scrutiny after an investigative journalist reported that an AI-generated edit fundamentally changed the meaning of his work. Instead of producing a neutral rewrite for narration, the AI introduced themes that were not present in the original investigation. The incident raised concerns about how generative models handle sensitive journalistic content.…
-
Concerns around Grok explicit images have pushed European regulators to take formal action against X. Authorities are examining whether the AI chatbot violated EU rules by generating sexually explicit and manipulated images, including non-consensual content. The case highlights growing pressure on major platforms to control the risks created by powerful generative AI systems. The investigation…
-
A new browser AI disabling tool gives users a way to remove built-in artificial intelligence features from Chrome, Microsoft Edge, and Firefox. The project targets browser integrations that automatically introduce AI tools, tracking components, and sponsored content. The developer created the tool for users who want a simpler browsing experience with fewer automated features and…
-
The ChatGPT Grokipedia source issue has sparked debate after users discovered that ChatGPT cited an AI-generated encyclopedia as a reference. The finding has renewed concerns about how large language models select sources and how AI-created content may influence factual accuracy. As AI systems increasingly shape how people access information, the credibility of their underlying sources…
-
Konni AI malware has surfaced in a targeted campaign aimed at blockchain engineers and developers. The activity marks a shift in focus for the threat group, which now targets technical professionals with access to high-value infrastructure and digital assets. The attackers rely on phishing delivery and AI-generated malware to compromise developer systems while avoiding traditional…
-
Fake ChatGPT VSCode extensions actively compromised developers by disguising spyware as AI-powered coding assistants. Attackers published these extensions in the Visual Studio Code Marketplace, where developers trusted them as legitimate productivity tools. The campaign exposed sensitive source code and development activity at scale. The incident shows how threat actors increasingly exploit interest in AI tools…