1. IRS Embraces AI to Enhance Taxpayer Services
The Internal Revenue Service (IRS) is now integrating Artificial Intelligence (AI) agents into its operations to tackle staff shortages and improve services for taxpayers. This strategic move aims to streamline processes, reduce the workload on existing employees, and ensure more efficient handling of taxpayer queries and issues. The adoption of AI is expected to significantly boost the IRS’s capacity to serve the public effectively amidst workforce challenges.
Read full story
2. ChatGPT Atlas Gets Vertical Tabs & Google Search
ChatGPT Atlas has rolled out significant updates, enhancing user experience with new features. Key additions include vertical tabs, which offer a more streamlined browsing interface. Users will now find Google integrated as the default search engine, making information retrieval quicker and more familiar. Furthermore, the updated version supports extension migration, allowing for a seamless transition of browser extensions. It also incorporates iCloud keychain integration, improving password management and security for Apple users. These updates aim to boost productivity and provide a more cohesive digital experience.
Read full story
3. Klay Vision Partners with Labels for Licensed AI Music Training
Klay Vision, an artificial intelligence (AI) music company, has announced significant licensing deals with several major record labels, including Warner Music Group. These agreements will allow Klay Vision to ethically train its Large Music Model (LMM) using a vast library of officially licensed music. This move is a landmark step in the AI music industry, ensuring that artists and copyright holders are compensated for the use of their intellectual property in developing AI-generated music. The partnerships aim to foster innovation in AI music creation while upholding ethical standards and ensuring fair compensation for creators.
Read full story
4. Honor Launches Robot Phone with AI Gimbal Camera
Honor has unveiled an innovative new smartphone, dubbed the ‘robot phone’, which features an advanced AI camera mounted on a mechanical gimbal arm. This unique camera system is designed to pop out from the back of the device, offering enhanced stability and dynamic shooting capabilities. The integrated artificial intelligence is expected to power sophisticated photography and videography functions, potentially introducing new ways for users to capture moments with exceptional clarity and smoothness. This launch marks a significant step in combining robotics with mobile technology.
Read full story
5. Google Launches Gemini 3.0: New AI Model Challenges Top LLMs
Google has officially launched Gemini 3.0, its latest iteration of the advanced large language model (LLM). The new version is currently undergoing rigorous testing, with claims suggesting it significantly outperforms other major LLMs available in the market. This release marks a significant step in AI development, as Google aims to set new benchmarks for performance and capability in artificial intelligence. Experts are keen to evaluate its real-world performance across various tasks and applications.
Read full story
6. AI’s Future: 900 Days Left, Says Stability AI CEO
Emad Mostaque, the CEO of Stability AI, recently made a significant statement, asserting that humanity has approximately 900 days to make critical decisions regarding artificial intelligence (AI). According to Mostaque, within this timeframe, the impact of AI on our world will become irreversible, fundamentally altering various aspects of society and technology. This urgent timeline highlights the need for immediate and comprehensive discussions on AI regulation, ethical guidelines, and its societal implications to shape a beneficial future.
Read full story
7. AI Fooled by Poetic Tricks: Study Reveals LLM Vulnerability
A recent study has shown that Large Language Models (LLMs) can be manipulated using cleverly crafted poems. Researchers successfully tricked these AI models in 62% of attempts by embedding harmful prompts within poetic structures. This finding highlights a significant vulnerability in current LLMs, suggesting that their ability to discern malicious intent can be bypassed through creative, subtle phrasing. The study underlines the need for more robust AI security measures to prevent such “poetic manipulation” from being exploited in real-world applications.
Read full story
8. Gemini 3 Refuses to Acknowledge Current Year, Claims Researcher Tricked It
Google’s advanced AI model, Gemini 3, recently displayed an unusual behaviour by refusing to recognise the current year as 2025. The AI reportedly accused a human researcher of attempting to trick it, indicating a potential glitch or an unexpected response to temporal queries. This incident highlights the complexities and challenges in developing and testing advanced artificial intelligence systems.
Read full story
9. NotebookLM Now Features Slide Deck Generation
NotebookLM has introduced a significant new feature: deck generation. This allows users to transform their source materials directly into customisable slide presentations. The feature is initially rolling out to Pro users, offering them an efficient way to convert notes and research into structured visual content. This enhancement aims to streamline the process of creating presentations from research data, making it quicker and more integrated within the NotebookLM environment.
Read full story
10. Memo Robot Learns Household Chores from Real Families
Memo, an innovative robot, has bypassed traditional lab training and is instead learning crucial household assistance skills directly from real-world environments. The robot’s impressive capabilities stem from its training on a massive dataset comprising 10 million real-world routines collected from 500 different families. This unique approach allows Memo to understand and adapt to various household tasks and preferences, making it highly effective for practical home assistance. This method ensures Memo’s learning is grounded in authentic human behaviour and diverse domestic settings.
Read full story
11. Ubisoft Tests AI Characters Responding to Player Voices
Ubisoft, a leading video game company, is currently experimenting with an innovative AI gaming technology. This new system allows non-player characters (NPCs) within the game to respond directly to players’ voice commands. This cutting-edge experiment aims to offer a more immersive and interactive gaming experience, enabling players to influence and personalise storylines through natural spoken interactions. This initiative could potentially revolutionise how open-world games and narratives are designed, providing unprecedented control and agency to gamers.
Read full story
12. Google Integrates Ads into AI Search Results
Google has initiated the display of sponsored content within its AI-powered search results, specifically in what it terms “AI Mode” responses. This move signifies the tech giant’s strategy to normalise the presence of advertisements within artificial intelligence generated search answers, potentially altering user experience and monetising new search functionalities. This integration aims to explore avenues for revenue generation through AI, similar to its traditional search engine model.
Read full story
13. Amazon AI Tool Now Public: AWS Open-Sources Agent SOPs
Amazon Web Services (AWS) has open-sourced its internally developed AI tool, Agent SOPs (Standard Operating Procedures). This innovative tool, designed to democratise AI expertise and automate repetitive tasks, was previously used within Amazon to streamline various operations. By making Agent SOPs publicly available, AWS aims to empower developers and businesses to leverage advanced AI capabilities for building more efficient natural language workflows for their own AI agents. This move is expected to foster broader adoption and innovation in the field of artificial intelligence.
Read full story
14. Louvre Heist: AI’s Weaknesses Exposed by Clever Thieves
A recent robbery at the Louvre Museum has shed light on critical vulnerabilities in artificial intelligence systems, specifically concerning their pattern recognition capabilities. The thieves cleverly exploited biases inherent in both human perception and AI algorithms, allowing them to evade detection. This incident underscores the need for more robust and nuanced AI development to counter sophisticated criminal tactics. By understanding these human psychology-based loopholes, we can improve AI’s ability to differentiate legitimate activity from deceptive actions, potentially preventing future security breaches.
Read full story
15. Google CEO: Advanced AI Agents Coming Soon
Google’s CEO, Sundar Pichai, has predicted that advanced Artificial Intelligence (AI) agents could be commonplace within a year. This significant announcement comes amidst revelations that Google has invested a massive $90 billion in AI infrastructure to support these technological advancements. Pichai’s statement highlights the rapid progress in AI development and Google’s commitment to leading the charge in this transformative field, anticipating a new era of highly capable AI systems.
Read full story
16. Spotting AI Music: How to Tell the Difference
Artificial intelligence (AI) is increasingly used in music creation, but experts say AI-generated tracks have distinct characteristics that make them identifiable. These tell-tale signs, often found in production techniques or performance nuances, can help music platforms and listeners differentiate between human-made and AI-composed music. This growing ability to detect AI music is leading platforms to develop methods for flagging such content.
Read full story
17. Caltech Unveils Drone-Launching Humanoid Robot
Researchers at Caltech have developed a groundbreaking humanoid robot that revolutionises aerial-ground interactions. This innovative robot is equipped with a integrated drone backpack, allowing it to autonomously carry, launch, and retrieve a drone. The robot boasts dual capabilities, functioning effectively both on the ground via driving mechanisms and in the air through its deployable drone. This development marks a significant step forward in robotics, enabling more versatile and dynamic operations in various environments.
Read full story
18. AI Uncovers New Lion Roar: A Breakthrough in Wildlife Research
Scientists have successfully employed Artificial Intelligence (AI) to identify a previously unknown intermediate roar sound made by lions. Analysing over 3,000 lion vocalisation recordings, the AI model achieved an impressive 95% accuracy in pinpointing this distinct roar, which bridges the gap between a gentle growl and a full-throated roar. This discovery is a significant advancement in understanding lion communication and behaviour, potentially aiding conservation efforts by providing deeper insights into their social structures and territory marking.
Read full story
