Header-template

 Global News

Subscribe our Newsletter

Subscribe to our mailing list to get the latest updates straight in your inbox.

TECH NEWS

Cisco Talos Warns Cybercriminals Weaponize LLMs Like WormGPT and FraudGPT

According to recent Cisco Talos research, cybercriminals are increasingly using large language models LLMs to improve hacking operations. Threat actor

According to recent Cisco Talos research, cybercriminals are increasingly using large language models LLMs to improve hacking operations. Threat actors are creating malicious programs, phishing emails, and hacking tutorials by employing uncensored or jailbroken LLMs, as well as by creating custom models like WormGPT, FraudGPT, and DarkGPT. These models use quick obfuscation techniques such as Base64 encoding, emojis, or "leet speak" to get beyond builtin safety safeguards. Some are integrated with thirdparty tools for automated exploitation, credential theft, and vulnerability scanning. Access to these tools has increased due to the growth of over 1.8 million models on websites like Hugging Face. This development represents a risky shift in the cybercrime ecosystem and lowers the bar for less experienced attackers. Read More..

B2B Tech News | 6 days ago          

Google Launches Doppl App to Let You Virtually Try On Clothes With AI

Doppl, an AIpowered tool developed by Google, allows users to virtually try on clothing by employing an animated digital version of themselves. Doppl,

Doppl, an AIpowered tool developed by Google, allows users to virtually try on clothing by employing an animated digital version of themselves. Doppl, which is accessible on Android and iOS in the United States, lets users submit a fullbody shot and pictures of their clothing from friends, retailers, or social media to see how they would appear in person. In order to increase realism, the app also produces AIgenerated videos that demonstrate how clothing moves. Favourite looks can be shared and saved by users. Although still in its early stages of development, Doppl expands upon Google Shoppings previous tryon tools and may occasionally provide inaccurate fit or detail information. At the moment, dresses, tops, and bottoms are supported, but shoes and accessories are not. Google intends to improve accuracy in response to user input. Read More..

B2B Tech News | 6 days ago          

Google Unveils Gemma 3n Multimodal AI That Runs Offline on Just 2GB RAM

Google has released Gemma 3n, a small, opensource AI model that can operate locally and offline on devices with as little as 2GB of RAM. Constructed u

Google has released Gemma 3n, a small, opensource AI model that can operate locally and offline on devices with as little as 2GB of RAM. Constructed using the MatFormer Matryoshka Transformer architecture, it generates text outputs in more than 140 languages and handles multimodal inputs, including text, graphics, audio, and video. There are two variations of the model E2B 5B parameters and E4B 8B parameters, which use PerLayer Embeddings PLE for elastic inference and optimised memory. Ideal for settings where privacy is a concern, Gemma 3n provides realtime applications like speechtotext, translation, and video analysis. Developers can use Google AI Studio, Kaggle, or Hugging Face to access it, and they can enter the Gemma 3n Impact Challenge to win a $150,000 prize fund. Read More..

B2B Tech News | 6 days ago          

WhatsApp Rolls Out AI Message Summaries With PrivacyPreserving Tech

WhatsApp has introduced AIpowered Message Summaries, which use Meta AI to distil unread messages into brief bullet points. In both group and individua

WhatsApp has introduced AIpowered Message Summaries, which use Meta AI to distil unread messages into brief bullet points. In both group and individual chats, the tool enables users to swiftly catch up on missed talks. It is presently only available in English in the United States, is optional, and is turned off by default. The user alone can see the summaries, which are created with Private Processing, a privacypreserving solution based on a Trusted Execution Environment TEE. This guarantees that messages and summaries are not accessible to WhatsApp or Meta. Through Advanced Chat Privacy settings, users can choose to enable the feature. The launch is in line with Metas goal of incorporating AI into commonplace apps while upholding robust privacy protections. Read More..

B2B Tech News | 17 days ago          

FileFix Exploit Uses Windows File Explorer to Execute Malicious Commands

FileFix is a new social engineering threat that targets Windows users by taking advantage of the address bar in File Explorer. A variation of the Clic

FileFix is a new social engineering threat that targets Windows users by taking advantage of the address bar in File Explorer. A variation of the ClickFix exploit, which deceived users into pasting malicious commands into PowerShell, FileFix was developed by researcher mr.d0x. Instead, FileFix tricks users into pasting a path into File Explorer by posing as a file location, which launches malicious code. Because it uses a reputable Windows interface, which reduces user suspicion, this technique is risky. The assault still presents a significant risk, even though Google Safe Browsing and Microsoft Defender SmartScreen may identify some payloads. Experts advise users to be wary of unusual popups or file requests and to refrain from pasting unknown paths. Read More..

B2B Tech News | 17 days ago          

Grok to Gain File Editor With Spreadsheet Support, Leaked Screenshots Reveal

Elon Musks xAI is apparently working on a Grok file editing function that will support spreadsheets and provide realtime AI support. A Grok side panel

Elon Musks xAI is apparently working on a Grok file editing function that will support spreadsheets and provide realtime AI support. A Grok side panel integrated in a spreadsheet enables users to speak with Grok while modifying papers, according to leaked photos supplied by app researcher Nima Owji. With the future Grok 3.5 model, the feature might make its debut and allow voice commands. This action puts Grok in a competitive position to take on productivity solutions like Google, Microsoft, and OpenAI however, specifics regarding additional file types are still unknown. If verified, Grok would make significant progress as it has only lately introduced web search, computer vision, and multilingual audio. Read More..

B2B Tech News | 17 days ago          

Google Launches Gemini Robotics OnDevice for Offline, RealTime Robot Control

Gemini Robotics OnDevice, a small voicelanguageaction VLA AI model that operates solely on robots without internet access, was introduced by Google. I

Gemini Robotics OnDevice, a small voicelanguageaction VLA AI model that operates solely on robots without internet access, was introduced by Google. It was created by DeepMind and allows biarm robots, such as ALOHA, Franka FR3, and Apollo humanoids, to use natural language commands to carry out complicated tasks, such as making belts or folding clothes. Designed for offline, lowlatency settings, the model requires less computing power and can learn new jobs with only 50–100 demos. Even with invisible objects or multistep instructions, it competes with cloudbased models in generalisation and dexterity despite its small size. Developers can test it using Googles MuJoCo simulator and access it through a Gemini Robotics SDK. This is a significant step towards autonomous, realtime robotics in both home and industrial environments. Read More..

B2B Tech News | 17 days ago          

TexttoSpeech Goes Mobile ElevenLabs Brings v3 Model to Android and iOS

ElevenLabs has brought its sophisticated voice production capabilities to smartphones with the release of its AI texttospeech mobile app for iOS and A

ElevenLabs has brought its sophisticated voice production capabilities to smartphones with the release of its AI texttospeech mobile app for iOS and Android. The software supports more than 70 languages, multispeaker conversations, and expressive audio tags like whisper and excited. It is powered by the Eleven v3 alpha architecture. In addition to exploring other alternatives designed for advertisements, narration, education, and social media, users can select from 19 basic voices. The app offers 10,000 free characters each month and shares credits on both mobile and web platforms. It facilitates rapid voiceover creation and smooth export to apps like CapCut and Instagram, making it ideal for producers, instructors, and marketers. ElevenLabs intends to enhance its mobilefirst content creation ecosystem in the future by incorporating AI agent tools and speechtotext capabilities. Read More..

B2B Tech News | 17 days ago          



Get The News to use newsletter to
stay on top of shifting Technology News Updates!!

...

Get Latest News To Your Inbox, Subscribe To Our Newsletter.


By signing up, you agree to our Privacy Notice and European users agree to the data transfer policy.



Take a break, Play Suduko!



Refresh Your Facts

1.  Global online sales are predicted to reach 22% of all retail sales in 2023.
   8731   

2.  Global tablet shipments are expected to hit 142 million by 2025, reports Statista.
   7241   

3.  Artificial Intelligence is expected to contribute $15.7 trillion to the economy by the year 2030.
   7594   

4.  Since the launch of the popular mobile game, Candy Crush, in 2012, players have spent a total of 73 billion hours (equivalent to 8.3 million years) playing this game.
   7698   

5.  Approximately 8 billion gadgets connect on the internet every day.
   8869   



Sharpen your mind and memory!

Play Mind Game



Your Opinion Matters, Vote Now!!

Where do you go to learn more about whats happening in your industry? Check all that apply.