Artificial Intelligence
Sber Presents its Neural Network’s New Version, Kandinsky 2.2

The new version of Sber’s generative model – Kandinsky 2.2 – can generate photorealistic images in higher resolutions and change the aspect ratio. Portraits now have significantly better quality. To update the neural network, Sber used a dataset of 1.5 bn text-image pairs, which is 300 mn more than had been used to train the previous version – Kandinsky 2.1.
The new version of the model can generate stickers to assemble fully-fledged sticker packs for Telegram. Kandinsky 2.2 bot can now change individual objects or elements on the image based on text descriptions while preserving the composition of the original image, thanks to ControlNet, an embedded dedicated managed change structural block.
Kandinsky 2.2 understands prompts in Russian and English, can paint in more than 20 different styles, blend several images, apply styles, generate images similar to input images, and support inpainting/outpainting. Alexander Vedyakhin, first deputy chairman of the Executive Board, Sberbank, “We continue to improve our neural network to create unique content. The new version provides more options to users and generates images of higher quality. Not only does our neural network try to imitate human creativity, but it can now create new artistic meanings and interpretations. As was the previous version, Kandinsky 2.2 is publicly available, so you can test it for free.”
You can try Kandinsky 2.2 in a Telegram bot, on the model’s promo page, at fusionbrain.ai, on ML Space in DataHub, a pre-trained model and dataset hub, in the Salute mobile app, and on Sber smart devices with the “Launch the Artist” voice command. Kandinsky 2.2 is a neural network developed and trained by the Sber AI team, which partnered up with the Artificial Intelligence Research Institute (AIRI), using a consolidated Sber AI and SberDevices dataset. The previous version, Kandinsky 2.1, had become the world’s fastest-growing neural network with one million active users engaged four days after launch.

Artificial Intelligence
Saudi and UAE Consumers Embrace AI-Powered Computers, Intel Survey Reveals

A new Intel survey on consumer attitudes toward AI-powered personal computers (AI PCs) reveals significant interest and adoption potential in Saudi Arabia and the UAE, particularly among younger demographics. The findings suggest a promising market for AI PCs, highlighting both excitement about the technology and areas where further education and reassurance are needed.
Key survey findings include:
- High Purchase Intent: Over 60 per cent of respondents would “definitely” upgrade to an AI PC, with another 29 per cent “probably” considering it, indicating a strong potential for rapid market growth.
- Strong Awareness, Low Usage: While 44 per cent of respondents know about AI PCs but have never used them, only 6 per cent currently own one. This gap presents a significant opportunity to demonstrate the practical benefits of AI PCs and convert awareness into adoption. This suggests a need for targeted marketing campaigns that showcase real-world applications.
- Feature Appeal: The most sought-after AI PC features include predictive text input, real-time language translation, and enhanced accessibility tools. These findings underscore the importance of focusing on user-centric AI applications that address everyday needs and improve productivity.
- Privacy Concerns Exist but Are Manageable: While 29 per cent of male and 28.4 per cent of female respondents are moderately concerned about data security, only a small percentage (6-7 per cent) are extremely concerned. This indicates that addressing privacy concerns proactively through transparent data handling practices and robust security features will be crucial for building trust and accelerating adoption.
- Time Spent on Routine Tasks Influences Attitudes: Those who spend over 20 hours per week on a PC are less likely to upgrade, perhaps due to satisfaction with their current setup or concerns about compatibility.
Conversely, mid-range users (10-19 hours per week) show the highest interest, suggesting that this group may be more receptive to the efficiency and productivity enhancements offered by AI PCs. These insights pave the way for tech developers and marketers looking to accelerate AI PC adoption in the region. The survey highlights the importance of focusing on user-centric features, addressing privacy concerns, and targeting specific user segments with tailored messaging.
Artificial Intelligence
HONOR and Google Cloud Deepen Partnership to Usher in New Era of Intelligent Devices

HONOR, a global leading AI device ecosystem company, showcased its continued collaboration pushing the boundaries of innovation in intelligent devices and user interaction with Google Cloud at Cloud Next 2025 last week. Eric Fang, Senior Director of Product Planning at HONOR, articulated the company’s visionary partnership with Google Cloud, has yielded many user-friendly innovations such as Magic Portal 2.0. Now, empowered by Google’s cutting-edge Gemini large language models (LLMs), this collaboration is entering a new era, focused on pioneering AI features that will redefine user experiences.
In March, HONOR unveiled its ambitious HONOR ALPHA PLAN, announcing its intention to become a leading global AI device ecosystem company. HONOR’s partnership with Google Cloud is a cornerstone of the ALPHA PLAN, providing access to cutting-edge AI infrastructure and expertise that will be instrumental in developing truly intelligent smartphones augmented with next-generation AI capabilities. HONOR is committed to breaking through technological boundaries in collaboration with partners like Google Cloud to co-create a new paradigm for AI devices. This partnership represents a crucial first step in realizing the ambitions of HONOR’s ALPHA PLAN.
Working closely with Google Cloud, HONOR will deeply integrate more AI functions into its smartphones, enriching the diversity of its imaging capabilities. Through close collaboration, the two companies are set to continue leading the way for innovation in both on-device and cloud-based AI, with all-new AI features that are poised to generate significant buzz among consumers.
During a panel discussion, Eric Fang shared insights into one of HONOR’s latest AI features, AI Outpainting. This innovative feature, part of HONOR’s AiMAGE suite, uses Google’s Imagen model to allow users to zoom out and expand their photos up to twice their original size. Users can reframe and rotate their photos after capturing them, reducing the need for careful pre-shot framing. “AI Outpainting is a game-changer,” commented Eric Fang. “It allows users to focus on capturing the decisive moment without worrying about perfect framing. The optimized generative fill ensures that the expanded images are seamlessly integrated, even in complex scenes like buildings and landscapes.”
Another highlight of HONOR’s collaboration with Google Cloud is the AI Eraser feature. This powerful tool allows users to easily remove unwanted objects from their photos with just a few taps. The integration of Imagen’s advanced AI technology ensures that the removed areas are filled with realistic and contextually appropriate content, minimizing the occurrence of AI hallucinations.
HONOR’s commitment to making AI accessible to everyone was a key theme during Cloud Next. By leveraging Google Cloud’s powerful infrastructure and cutting-edge AI capabilities, HONOR aims to empower users globally with innovative features that enhance their everyday experiences. To this end, HONOR is committed to making flagship-level AI features available on its midrange devices wherever possible, for example with the recent launch of the HONOR 400 Lite in Europe on April 7th, which includes the AI Outpainting, AI Eraser and AI Upscale features.
“Democratizing AI is at the core of our mission,” said Eric Fang. “We want to make advanced AI features available to everyone, unlocking the full potential of intelligent technology for a smarter future.” HONOR’s deep collaboration with Google Cloud represents a significant milestone in its unwavering dedication to delivering leading-edge technology and exceptional user experiences. Beyond the specific AI features unveiled at Google Cloud Next 2025, the partnership underscores HONOR’s proactive and future-oriented approach to AI integration.
By working closely with global technology leaders like Google Cloud, HONOR is reinforcing its commitment to pushing the boundaries of innovation and providing users with intuitive, powerful, and truly intelligent devices that anticipate their needs. This collaboration is a testament to HONOR’s confidence in the transformative power of AI and its strategic vision for the future of mobile technology. Eric Fang also teased upcoming products that will integrate even more advanced AI features, promising exciting new experiences for users. Fang announced that HONOR has collaborated deeply with Google Cloud on its upcoming new products, set to launch in May 2025.
Leveraging Veo, Google’s state-of-the-art video generation, HONOR will introduce co-developed experiences that accurately represent motion with enhanced realism. This integration has the potential to interpret user commands and generate videos that represent their creative vision, unlocking new possibilities for artistic expression. Furthermore, to significantly enhance user productivity, HONOR will feature more advanced office tools on these upcoming devices, co-developed with Google Gemini.
These tools are designed to streamline workflows and empower users with intelligent assistance for everyday tasks. HONOR’s collaboration with Google Cloud will continue to drive innovation in the combination of device and cloud AI, setting new standards for intelligent devices. Both parties are also continuously co-developing futuristic AI features to create truly intelligent smartphones, serving users and making an impact in global markets.
Artificial Intelligence
HONOR Brings AI Deepfake Detection to the Middle East on HONOR Magic7 Pro

With the rise of AI technology, the threat of scams and digital manipulation is growing. Deepfakes—hyper-realistic but entirely fake images and videos—are becoming more common, making it harder to trust what we see online. As AI continues to evolve, so do the risks to our privacy, security, and even personal relationships. As a human-centric brand putting security first, HONOR believes that user’s privacy and safety should always come first. The brand has invested significantly in developing on-device AI Deepfake Detection technology in response to the safety risks associated with deepfakes.
Now, HONOR is officially launching this AI Deepfake Detection feature in the Middle East, bringing advanced on-device protection to users across the region. HONOR is the first smartphone brand to introduce on-device AI Deepfake Detection technology to smartphones. The AI Deepfake Detection feature works during live video calls, providing users with real-time protection and peace of mind while communicating face-to-face.
The feature debuts exclusively on the HONOR Magic7 Pro, with plans to expand to more HONOR devices in the near future. This feature is designed to detect manipulated images and videos in real time, helping users identify deepfakes across popular apps like WhatsApp, Facebook, Botim, WeChat, Telegram, and more.
When developing HONOR’s on-device AI Deepfake Detection feature, the brand’s R&D team prepared a significant dataset comprising millions of videos and pictures, which was then used to train an on-device deepfake-detection AI model. HONOR also used AI to closely inspect tiny details in each pixel that the naked eye could miss. This commitment to safety and security is part of the brand’s broader strategy to integrate AI deeply into the user experience.
HONOR previously introduced its Four-Layer AI Architecture, highlighting its strategic commitment to not only integrating AI into MagicOS but also ensuring it is done with a strong emphasis on security. This Four-Layer Architecture comprises distinct layers. At the base layer, Cross-device and Cross-OS AI form the foundation of an open ecosystem, which allows the sharing of computing power and services among devices and operating systems. Building upon this foundation, the Platform-level AI layer enables a personalized operating system, allowing intent-based human-computer interaction and personalized resource allocation.
At the third layer, App-level AI is poised to introduce a wave of innovative, generative AI applications that will revolutionize user experiences. Lastly, at the top, the Interface to Cloud-AI services layer provides users with easy access to massive cloud services while prioritizing privacy protection, creating a truly holistic and future-forward AI experience.
With the launch of AI Deepfake Detection in the Middle East, HONOR is taking a bold step toward a safer digital future. As deepfakes and AI-driven scams continue to evolve, HONOR remains committed to staying ahead of the curve—offering users intelligent, on-device protection that puts privacy first.
You must be logged in to post a comment Login