V-Help
We are helping with: servers and virtualization
IT support for businesses: on-site and remote help, clear estimates and reports.
32
services in catalog
2–4 h
typical response time
Moscow
and region, on-site
How we work
Request
Call, form, or email — we capture the task.
Diagnostics
Remote or on-site review of your infrastructure.
Resolution
We fix the issue and provide a summary.
Pick a block to open the matching section on Services
Latest IT and technology news
Page 5 of 244 · 2921 total
Artificial intelligenceAfter discussing fish recipes, Google Home’s AI assistant unexpectedly renamed a user to 'Fish Lemon,' displaying the odd name in reminders, smart displays, and even direct name queries. Attempts to restore the original name have failed, as the system stubbornly reverts to the playful nickname. Such incidents are not isolated; AI has previously assigned fictional names to people in camera frames.
Read more
Artificial intelligenceTechCrunch Disrupt 2026 expands its AI-focused program by launching a new Real World AI stage, dedicated to bridging digital and physical technologies. The stage will cover autonomous systems, robotics, and even projects aimed at reviving extinct species. Confirmed speakers include representatives from Nvidia, Shield AI, Colossal Biosciences, and startups specializing in edge computing and industrial AI. The event will take place from October 13 to 15 in San Francisco, gathering over 10,000 participants from tech, startups, and venture capital sectors.
Read more
HardwareGooglebooks laptops running on Android-based Aluminum OS will introduce 'Device Intelligence,' an AI-powered feature that automatically sorts files into thematic collections. For example, all documents related to an upcoming trip will be grouped into a single folder. Access to these folders will be available via desktop widgets without opening the Files app. The feature is disabled by default, and all data remains local—never sent to Google servers or used for AI training.
Read more
SecurityCybersecurity experts have detected mass attacks targeting the CVE-2026-9586 vulnerability in Sangoma Switchvox, a corporate VoIP platform. This flaw allows threat actors to execute arbitrary code on vulnerable systems without authentication. According to Horizon3 researchers, most internet-exposed Switchvox instances have already been compromised or remain at high risk. Administrators are urged to update the software to the latest version and scan systems for signs of compromise.
Read more
Artificial intelligenceGoogle continues to rapidly advance its AI models, releasing the third version of Gemini Flash 3.8 in just six weeks. This update follows a temporary pause in the development of the Pro model, highlighting the company's focus on optimizing lighter, faster solutions. The new version aims to enhance performance and algorithmic efficiency while maintaining market competitiveness in AI technologies. Developers note that Gemini Flash is designed for tasks requiring high-speed data processing with minimal resource consumption.
Read more
Artificial intelligenceDeveloper Christopher tested how AI-driven hiring systems would handle another AI as a candidate. Instead of a traditional interview, he used ChatGPT to respond to questions from an automated interviewer. The experiment highlighted both the strengths and limitations of modern HR technologies, prompting reflection on the future of AI-to-AI interactions in corporate environments.
Read more
Artificial intelligenceArtificial intelligence has made remarkable strides in solving complex tasks, including the renowned New York Times Connections puzzles. Within months, top models achieved near-perfect accuracy, whereas previously they solved only 18% of challenges. These tests not only highlight AI's progress but also expose its limitations, where humans still outperform machines.
Read more
SecurityMicrosoft has announced plans to expand the Memory Integrity feature in Windows 11, which will be automatically enabled on more compatible PCs beginning in October 2026. The update will roll out via standard quality patches and strengthen kernel protection against attacks. Devices where users previously disabled this option will remain unaffected. The feature leverages Virtualization-based Security (VBS) to isolate kernel code integrity checks, preventing the execution of malicious software and untrusted drivers.
Read more
Artificial intelligenceThe Russian startup Mostik has introduced a groundbreaking approach to AI interaction, eliminating text-based intermediaries. Developed by a team of mathematicians, this method enables AI models to exchange information directly through structured data, significantly reducing processing time and computational load. The technology holds potential for machine learning systems requiring high-speed responses and minimal latency.
Read more
HardwareXiaomi has set September 7 as the date for its flagship device autumn launch event. The presentation will introduce the foldable smartphone Xiaomi 18 Fold, the first device powered by the new flagship Xring O3 processor with AI support. Additionally, the Xiaomi Pad 9 Pro Max tablet and three Sky Nomad electric vehicle models—N70 Pro, N70 Max, and N90 Max—will be unveiled. This marks the first official launch date for the electric vehicle lineup following the start of pre-orders in July.
Read more
Artificial intelligenceZ.ai, the international brand of Chinese AI model developer Zhipu, has launched its official store on Tmall, marking the first time AI service subscriptions are available to mass audiences through a major e-commerce platform. Users can now purchase AI tool access plans—Lite, Pro, and Max—via Taobao’s standard checkout process. Zhipu notes that subscription discounts may be offered during major sales events, making AI services more accessible.
Read more
SecurityMicrosoft has confirmed a critical issue in its Defender for Office 365 cybersecurity solution, where legitimate Google search result links are being incorrectly blocked as malicious. Users encounter security warnings when attempting to access these links, disrupting workflows. The incident, tracked under ID MO1465962, was detected at 10:30 UTC. Directly copying and pasting links into browsers does not bypass the issue, as the error stems from incorrect threat classification in the system.
Read more
