APIs

AppWizard
August 23, 2025
Google is developing a new settings page called "Hearing Wellness" for Pixel users, which will integrate Android's Sound Exposure and Hearing Health options. This feature will allow devices to automatically adjust volume levels when they exceed safe thresholds, with users having the option to enable or disable this automatic volume reduction. The initiative responds to user feedback regarding dissatisfaction with previous automatic adjustments for regulatory compliance. The new interface includes toggles for Sound Exposure notifications and Hearing Health. Additionally, Android 16 QPR2 Beta 1 has been released for enrolled Pixel devices, introducing enhancements like an expanded dark theme and auto-themed icons, as well as new developer APIs.
AppWizard
August 21, 2025
Google has released Android 16 QPR2 Beta 1 following the Pixel 10 series launch, aimed at developers. The update includes an expanded dark theme, auto-themed icons, and developer-focused APIs. The beta will run until Q4 2025, with QPR1 launching next month featuring Material 3 updates. The beta is available for compatible Pixel devices enrolled in the previous Android 16 beta, and new users can join via google.com/android/beta. The build number for the beta is BP41.250725.006, which includes the August security patch. Users must opt out of the beta program to transition to the stable Android 16 QPR1 update next month, which will introduce Material 3 design and additional features.
AppWizard
August 7, 2025
A glitch in the YouTube app for Android prevents users from adjusting playback speeds, leaving them stuck at the default 1x speed. This issue has been reported by users globally and mirrors a similar problem from the previous year, indicating a recurring vulnerability in YouTube’s Android codebase. YouTube has acknowledged the issue and is investigating it, with speculation that it may be linked to recent Android OS updates or changes to internal APIs. The bug affects both stable version 20.28.39 and beta version 20.29.39, while YouTube Music's podcast controls remain unaffected. Users have shared workarounds, such as uninstalling app updates to revert to a functional version, although this may pose security risks. The incident raises concerns about Google’s quality assurance processes and the balance between innovation and reliability in app development.
Winsage
August 6, 2025
Microsoft has integrated OpenAI’s GPT model, gpt-oss-20b, into its Windows AI Foundry platform, allowing Windows 11 users to access advanced AI capabilities from their desktops. The model is optimized for agentic tasks and is compatible with consumer PCs and laptops with a minimum of 16GB of VRAM. gpt-oss-20b is a text-only model with a 53% hallucination rate on OpenAI’s PersonQA benchmark. Microsoft plans to extend its availability to macOS and other devices, and both gpt-oss-20b and its predecessor will be accessible through Microsoft’s Azure AI Foundry and Amazon’s AWS.
Winsage
August 6, 2025
AWS has introduced two new OpenAI models with open weights, the gpt-oss-120b and gpt-oss-20b, available through Amazon Bedrock and Amazon SageMaker JumpStart. These models are designed for text generation and reasoning tasks, excelling in coding, scientific analysis, and mathematical reasoning, with performance comparable to leading alternatives. They support a context window of 128K and adjustable reasoning levels (low, medium, high). Users can access these models via an OpenAI-compatible endpoint in Bedrock, utilizing the OpenAI SDK or Bedrock APIs for integration. To access the models in Amazon Bedrock, users must request access through the console and can evaluate them using the Chat/Test playground. The process for using the OpenAI SDK involves configuring the API endpoint and authentication with an Amazon Bedrock API key. Users can build AI agents using frameworks that support the Amazon Bedrock API or OpenAI API, with deployment facilitated by Amazon Bedrock AgentCore. The OpenAI models are available in the US West (Oregon) region through Amazon Bedrock and in the US East (Ohio, N. Virginia) and Asia Pacific (Mumbai, Tokyo) regions via SageMaker JumpStart. Each model provides full chain-of-thought output capabilities for enhanced interpretability. They can be modified and customized to meet specific needs, and security measures are in place to ensure robust evaluation processes. Compatibility with the standard GPT-4 tokenizer is also included. For cost details, users can refer to the Amazon Bedrock and SageMaker AI pricing pages.
Winsage
August 5, 2025
A new variant of the RoKRAT malware, attributed to North Korea's APT37 group, utilizes advanced techniques such as steganography to hide malicious code within JPEG image files, complicating detection efforts. This malware is primarily distributed in South Korea through compressed archives containing Windows shortcut files that lead to a multi-stage infection process. The process involves executing PowerShell commands to decrypt and run the malware, which can inject itself into trusted Windows processes like mspaint.exe and notepad.exe, leaving minimal forensic traces. The malware also exfiltrates sensitive information using legitimate cloud APIs, making attribution difficult. APT37 has demonstrated adaptability by changing its injection targets and camouflaging its development artifacts, highlighting the need for advanced Endpoint Detection and Response (EDR) solutions and proactive security measures.
Winsage
August 4, 2025
Specialists at the Genians Security Center have identified a new version of the RoKRAT malware linked to the North Korean APT37 group. This version uses steganography to hide its code in JPEG images, allowing it to bypass antivirus systems. The infection begins with a malicious .LNK link in a ZIP archive, which contains a large .LNK file that misleads users. The malware employs various encrypted components, including shellcode, PowerShell scripts, and batch files. Upon execution, PowerShell decrypts the shellcode using a XOR operation, and the malware injects itself into legitimate Windows processes without leaving traces on the disk. The RoKRAT loader is embedded in a JPEG image hosted on Dropbox, and it uses a double XOR transformation to extract the shellcode. The malware is activated through sideloading techniques using legitimate utilities and downloads from cloud platforms. RoKRAT can collect data, take screenshots, and transmit them to external servers. Recent samples have targeted “notepad.exe” for code injection, indicating ongoing development. Endpoint detection and response (EDR) systems are essential for monitoring unusual activities and protecting against these sophisticated attacks, as traditional defenses are inadequate.
Winsage
August 4, 2025
Security researchers at Genians Security Center discovered a new variant of the RoKRAT malware linked to the North Korean APT37 threat group. This malware uses steganography to hide malicious payloads within JPEG files, allowing it to evade traditional antivirus detection. It is typically distributed through malicious shortcut files within ZIP archives, often disguised as legitimate documents. The malware employs a two-stage encrypted shellcode injection method, utilizing PowerShell and batch scripts to execute its payloads in memory. It collects system information, documents, and screenshots, exfiltrating data via compromised cloud APIs. The command and control accounts associated with the malware are linked to Russian email services. Variants of RoKRAT have evolved to include different injection methods and reference specific PDB paths. Indicators of compromise include various MD5 hashes associated with the malware.
Search