NerdNews

OpenAI's ChatGPT is expected to hit 700 million weekly active users, with a 4x growth since last year, driven by its upgraded image-generation feature and increasing subscriber base.

Tesla's board of directors has announced a new compensation package for CEO Elon Musk worth around $29 billion in shares, citing the 'ever-intensifying AI talent war' as a reason. The package includes 96 million shares that will vest in two years if Musk serves continuously in a senior leadership role and holds the stock for five years.

Zero day exploitation increased by 46% year-over-year in the first six months of 2025, with products from 27 vendors impacted. Microsoft and Google were among the most affected, with 30% and 11% of zero day exploits respectively. The Forescout report also found that ransomware attacks grew 36% year-over-year, with 3649 documented attacks in H1 2025.

Google's Med-Gemini AI model made a critical error by inventing a non-existent body part, 'basilar ganglia', which is a combination of two real terms. Experts warn this could have serious consequences if doctors don't notice the mistake. The incident highlights the risks of relying on AI in medicine and the need for rigorous testing and validation.

Anysphere CEO Michael Truell talks about Cursor, an automated programming platform that integrates with generative AI models to help write code. He discusses the potential of AI in programming, the challenges of automating code, and the future of programming. Truell also shares his thoughts on the adoption of AI in the tech industry and the potential for AI to replace certain jobs.

Forrester warns that major enterprise software vendors like Oracle, SAP, and Salesforce are using AI to end discounting and push high-margin products, increasing vendor lock-in and strategic risk for customers.

Cloud intrusions have increased by 136% in the first half of 2025 compared to all of 2024, with Chinese-nexus actors exploiting cloud environments and threat actors using manual navigation to bypass legacy detection tools.

AI agents are transforming security architecture by introducing autonomous decision-making, creating new challenges in identity, access, and trust. To address these challenges, organizations must adopt new security strategies, including least privilege, proper identity architecture, and building trust through frameworks. The emergence of Model Context Protocol (MCP) servers and agent-to-agent authentication will require new approaches to security and compliance.

A customer data platform (CDP) is a software application that supports marketing and customer experience use cases by unifying a company's customer data. Gartner recommends evaluating CDP providers' capabilities, including data collection, customer profile unification, integrations, segmentation, experimentation, and data science and privacy. CDPs can provide significant value to marketing and adjacent functions, enabling business users to activate rich data and insights across channels, devices, and enterprise data applications.

The article showcases 12 examples of people who have created powerful internal tools by prompting AI, replacing expensive software and automating workflows without writing code. These tools have saved six figures, launched products faster, and delivered meaningful outcomes. AI has made it possible for anyone to build custom software tools, regardless of their coding abilities.

Google's AI-powered bug hunter, Big Sleep, has found and reported 20 security vulnerabilities in popular open source software, including FFmpeg and ImageMagick. The tool, developed by DeepMind and Project Zero, uses LLMs to look for and find vulnerabilities, with a human expert verifying the reports before they are submitted.

Grok Imagine, xAI's new AI image and video generator, allows users to create content, including NSFW images and videos, with a 'spicy mode' and generates content in seconds from text prompts. The model is still in early days and has limitations, but it's impressive and aims to compete with Google DeepMind and OpenAI.

Lyft and China's Baidu are partnering to bring robotaxis to Europe next year. The companies plan to launch services in Germany and the UK in 2026, pending regulatory approval. Baidu's Apollo Go autonomous vehicles will be integrated into Lyft's ride-hailing app.

A 50% increase in web traffic to AI-related websites has been reported, with 80% of GenAI usage occurring within browsers. This shift raises security concerns, including a 130% increase in zero-hour phishing attacks driven by AI.

Foxconn is converting its US electric vehicle factory to produce AI servers due to low demand and overcapacity in the EV market. This move aligns with the company's AI ambitions and comes as Tesla's CEO Elon Musk receives a $29 billion award for shifting the company's focus to AI, robotics, and related services.

The Pwn2Own competition is offering a $1 million bounty for a zero-click WhatsApp exploit. The competition, which focuses on consumer products, will take place in Cork from October 21 to 24. Researchers will have the chance to hack into various devices, including mobile phones, smart home devices, and wearables. The goal is to incentivize security researchers to find exploits and responsibly disclose them to vendors.

A mountain rescue team used AI to analyze drone footage and locate the body of a missing hiker in the Italian Alps. The AI system identified the hiker's helmet in just one afternoon, saving time and resources in the recovery operation.

Ransomware gangs have joined attacks targeting a Microsoft SharePoint vulnerability chain, exploiting flaws to breach at least 148 organizations worldwide. The 4L4MD4R ransomware variant has been detected, demanding 0.005 Bitcoin and encrypting files on compromised systems.

A patient with ALS has demonstrated the ability to control an iPad using only their thoughts, thanks to a prototype developed by Synchron using Apple's Brain-Computer Interface Human Interface Device (BCI HID) protocol. The patient, Mark Jackson, is able to launch apps and fully use them despite being unable to use his hands. This breakthrough has the potential to greatly improve the lives of people with motor disorders.

Microsoft admits it can't guarantee data sovereignty, sparking concerns about cloud data security and the ability of hyperscalers to protect nation's data from foreign powers. This raises questions about the future of cloud computing and data storage.