IT News Review by Control F5 Software: More and more users are developing emotional relationships with AI chatbots

More and more users are developing emotional relationships with AI chatbots

An increasing number of people are using chatbots like ChatGPT or Claude not only for productivity but also for personal conversations and emotional support, according to an Associated Press analysis. The phenomenon is drawing the attention of researchers and ethics specialists who are trying to understand how people interact with conversational systems.

AI models are capable of maintaining coherent dialogues, adjusting their tone, and providing empathetic responses, which can create the impression of a personal relationship. However, specialists emphasize that these systems do not understand emotions in a human sense and cannot replace social relationships or the support provided by professionals.

For AI product developers, the situation raises new challenges regarding the design of interactions, user protection, and the limits that conversational assistants should respect. As technology becomes more convincing, the responsibility for how it is used increases significantly.

The subject highlights that the evolution of AI not only brings technological changes but also social, psychological, and ethical effects that will influence the development of future digital products.

Financial regulators accelerate AI adoption to combat new cyber risks

Financial supervisory authorities worldwide are intensifying investments in artificial intelligence to respond to a new generation of cyber threats. As financial institutions increasingly use AI for automation, analysis, and decision-making, attackers are using the same technologies to identify vulnerabilities and launch faster and more sophisticated attacks.

According to representatives of the Swiss authority FINMA and the international organization IOSCO, which brings together authorities overseeing about 95% of global financial markets, the solution is to develop SupTech tools – AI-based technologies for monitoring and automatic risk identification. Meanwhile, American and European authorities are analyzing how banks control the use of AI, focusing on governance, data access, third-party providers, and emergency stop mechanisms for automated systems.

This evolution marks a significant change in how technology regulation is viewed. It is no longer sufficient to simply monitor how companies use artificial intelligence. Authorities are developing their own AI tools to keep pace with the speed of attacks and the complexity of modern digital infrastructures, especially in areas like financial services and digital assets.

More and more officials warn that autonomous AI systems may require new regulatory rules, different from the current framework. Among the measures being analyzed are automated transaction limiting mechanisms, rapid recovery procedures in case of incidents, and additional requirements for monitoring AI models used in critical operations.

Study: Most restaurants are missing from AI-generated recommendations

A new study shows that the emergence of AI-based search engines is rapidly changing how users discover restaurants, but most businesses are not visible in these recommendations. Research conducted by Local Falcon shows that nearly 75% of restaurants do not appear at all in AI-generated responses when users search for places to eat.

The difference between the results provided by Google Maps and those generated by AI is significant. While Google Maps displays a wider variety of locations, AI models tend to recommend only a very limited number of restaurants, with the top-ranked ones concentrating most of the visibility. Thus, a small number of brands end up dominating the recommendations, while most restaurants remain practically invisible to users.

For the industry, the conclusion is relevant beyond the HoReCa sector. As users begin to use AI assistants for local searches and recommendations, optimization exclusively for traditional search engines is no longer sufficient. Visibility in AI ecosystems becomes a new objective for digital strategies, and structured data, reviews, and online reputation gain increasing importance.

The study also highlights a broader trend: AI does not provide dozens of results but only a few recommendations deemed relevant. This fundamentally changes the competition for digital visibility and can influence how companies build their online presence in the coming years.

Zhipu AI bets on open source to compete with OpenAI and Anthropic

The Chinese company Zhipu AI is accelerating the development of open-source models and aims to become one of the main competitors to American leaders in artificial intelligence. The strategy comes at a time when the global AI market is marked by massive investments, and differentiation is increasingly made through ecosystems and technology accessibility, not just through model performance.

Company representatives argue that open-source models allow for faster adoption in the enterprise environment and provide developers with the necessary flexibility for integration into their own applications. Zhipu AI is simultaneously investing in expanding infrastructure and developing multimodal models capable of processing text, images, and other types of content, trying to reduce the gap with OpenAI and Anthropic.

The competition between proprietary models and open-source ones continues to redefine the AI market. For companies, the choice is no longer based solely on performance but also on costs, customization possibilities, control over data, and independence from a single provider. These criteria are becoming increasingly important as AI is integrated into critical applications and processes.

The evolution also confirms the trend of fragmentation in the global market. Alongside established players from the United States, more and more companies from Asia are investing in developing their own models, leading to accelerated competition both technologically and geopolitically.

AI users are starting to focus on efficiency, not just model consumption

The artificial intelligence industry is entering a new stage, where success is no longer measured solely by the number of queries or time spent in applications. According to a CNBC analysis, users and companies are beginning to use AI models in a much more efficient way, focusing on concrete results and automating high-value processes.

This change also modifies how AI providers build their business strategies. OpenAI, Anthropic, and other major players are trying to develop products that generate value by integrating into the workflows of organizations, not just through occasional use of chatbots. The focus is shifting towards AI agents, productivity tools, and solutions dedicated to the enterprise environment.

For organizations, this evolution means that the return on investment in AI will increasingly be evaluated based on its impact on productivity, quality, and execution speed. The number of users or conversations becomes a secondary indicator compared to time savings and operational efficiency achieved.

The change reflects the maturation of the entire industry. After a period of experimentation and rapid adoption, companies are beginning to seek sustainable and measurable uses, which will influence the development of future AI products.

YouTube Shorts introduces double-speed playback

YouTube continues to adapt the Shorts platform to user behavior and introduces the ability to play videos at 2x speed. The new feature is available by long-pressing the screen during playback and complements existing options for rapid content consumption.

The update aims to give users more control over how they watch short videos. The feature is already familiar on other platforms and responds to the preferences of an audience that consumes very large volumes of video content daily.

For content creators, the change may influence how video materials are built. The pace of narration, the duration of scenes, and the presentation style must be adapted to remain easy to follow even at accelerated playback speeds.

The update confirms the intense competition between short video platforms, where every optimization of the user experience can influence the time spent in the application and the level of audience engagement.

Adobe acquires Topaz Labs to expand its AI photo and video processing capabilities

Adobe has announced the acquisition of Topaz Labs, a company known for its AI-based tools dedicated to enhancing images and videos. The solutions developed by Topaz are widely used for noise reduction, resolution enhancement, and restoration of photos and video recordings.

The integration of Topaz Labs technology into the Adobe ecosystem aims to strengthen the AI features available in applications like Photoshop, Lightroom, and Premiere Pro. The company thus continues its strategy of integrating artificial intelligence into creative processes, reducing the time required for editing and automating complex operations.

For professionals in the digital creative field, the acquisition promises access to more powerful tools directly in existing workflows. At the same time, Adobe is strengthening its position in a market where AI is becoming an important differentiator for photo and video editing applications.

The transaction reflects the broader trend of consolidation in the software industry, where major platforms prefer to integrate already validated AI technologies on the market rather than develop them exclusively in-house.

Facebook launches an AI assistant dedicated to content creators

Meta has begun rolling out a new AI application aimed at content creators on Facebook. The new assistant is designed to support the process of creating, planning, and optimizing content, providing suggestions and recommendations tailored to each creator's activity.

The application can analyze the performance of published content and propose ideas for new materials, text variations, titles, or publishing strategies. The goal is to reduce the time spent on repetitive activities and facilitate the management of online communities.

For creators and companies using Facebook as their main communication channel, the integration of AI into the platform's tools can accelerate content production and simplify management processes. At the same time, differentiation will continue to depend on creativity and the ability to build an authentic relationship with the audience.

The launch is part of Meta's strategy to integrate artificial intelligence into all its products, making AI a standard component of the user experience for individuals and companies.

OpenAI develops its first custom AI chip together with Broadcom

OpenAI has unveiled its first custom chip for artificial intelligence, developed in collaboration with Broadcom. The initiative aims to reduce dependence on traditional AI processor suppliers and optimize costs associated with training and running large models.

The processor is specifically designed for the requirements of OpenAI models and will complement the existing GPU-based infrastructure. The development of dedicated hardware gives the company greater control over performance, energy consumption, and operating costs.

This move reflects the direction taken by more and more companies in the AI field, which are investing in their own processors to meet the growing demand for computing power. Hardware is thus becoming a competitive advantage as important as software models.

As AI continues to expand, the competition between infrastructure providers and model developers will increasingly include the design of their own specialized chips capable of supporting huge processing volumes with more efficient energy consumption.

Claude learns about organizations by analyzing conversations in Slack

Anthropic is expanding the capabilities of the Claude platform through a new feature that allows the AI model to better understand the context of an organization using the information available in Slack. The functionality, called Claude Tag, can identify terms, projects, teams, and internal processes, providing more relevant responses for employees.

The system analyzes messages and documentation available in the Slack channels that users already have access to, without extending existing permissions. The goal is to build an organizational memory that allows the AI to respond based on the specifics of the company, reducing the time spent searching for information and explaining context in each conversation.

For companies, this approach represents an important step towards AI assistants integrated into daily workflows. Instead of functioning as a generalist chatbot, the model can understand internal projects, specific terminology, and relationships between teams, becoming a more useful tool for operational activities.

The trend confirms the direction in which enterprise artificial intelligence is evolving. The value of models is no longer determined solely by general knowledge but also by the ability to understand the context of each organization and to use available information in a safe and controlled manner.

New information about the future iPhone Ultra Foldable

New information from the supply chain suggests that Apple is continuing to develop its first foldable iPhone, unofficially named iPhone Ultra Foldable. According to reports, the company is working on a device with a large internal screen and a redesigned hinge to reduce the visibility of the fold.

Apple seems to prioritize reliability and user experience ahead of the product launch. Sources indicate that the company is testing new materials for the screen and folding mechanisms that can withstand a large number of usage cycles, avoiding issues encountered with the first generations of foldable phones on the market.

If the information is confirmed, Apple's entry into this segment could accelerate the adoption of foldable devices and intensify competition with manufacturers like Samsung, Huawei, or Honor. At the same time, the iOS ecosystem would need to be adapted for specific use scenarios of flexible screens and advanced multitasking.

Although the company has not officially confirmed the product's launch, interest in this category continues to grow, and the market expects Apple to enter the segment only when it believes the experience offered meets its quality standards.

Australia tightens the ban on children's access to social networks

The Australian government has announced additional measures to enforce the ban on social media use by children below the legally established age limit. The new amendments provide for significantly higher financial penalties for technology companies that do not comply with user age verification obligations.

Authorities argue that digital platforms must demonstrate that they are making real efforts to prevent minors' access and cannot transfer responsibility solely to users or parents. Fines are increasing significantly compared to the current level and aim to encourage the implementation of more effective technical mechanisms.

For major social platforms, the new rules imply additional investments in identity verification systems, data protection, and automated mechanisms for detecting underage users. At the same time, the measures bring back into focus the debate about the balance between child protection and user privacy.

Australia thus continues to adopt one of the strictest approaches globally regarding the regulation of social platforms, a trend closely monitored by other states considering similar measures.

Austria seeks to attract a European Anthropic center

Austria is conducting a campaign to host one of the future European centers of Anthropic, the developer of the Claude model. The government believes that the presence of an important company in the field of artificial intelligence would strengthen the country's position as a technology and research hub in Europe.

The initiative includes discussions about facilities for investments, collaborations with universities, and access to research infrastructure. The competition is intense, with several European states trying to attract similar investments in the context of the accelerated growth of the AI industry.

For the European technology ecosystem, the establishment of AI development centers represents more than just direct investments. They contribute to the development of local skills, attract specialists, and stimulate collaboration between the academic environment and the private sector.

The growing interest of European states in such investments shows that AI infrastructure is beginning to be viewed as a strategic element, comparable to investments in data centers, semiconductors, or advanced research.

Google limits the use of Gemini models by Meta

Google has introduced restrictions on how Meta can use certain Gemini models, according to a report published by the Financial Times. The decision reflects the growing tensions between major companies in the field of artificial intelligence, which are simultaneously competing in the development of models, infrastructure, and commercial products.

Information indicates that Google wants to more strictly control the use of its technology by direct competitors, especially in situations where AI models could contribute to the development of competing products. The relationships between model providers and major platforms are thus becoming increasingly complex.

For companies building AI applications, the situation highlights the importance of diversifying suppliers and avoiding dependence on a single technological ecosystem. Access to high-performance models can become a competitive advantage but also a vulnerability if commercial or technical terms change.

As the AI market matures, competition extends beyond model performance and includes control over infrastructure, data, and commercial relationships among key industry players.

Kyivstar and the Ukrainian government collaborate to develop digital infrastructure

The telecom operator Kyivstar and the VEON group have signed a partnership with the Ministry of Economy of Ukraine and other institutions to develop digital infrastructure and support the country's economic reconstruction process.

The project aims to accelerate digital transformation through investments in connectivity, digital services, and modern infrastructure capable of supporting public administration and the private sector. The initiative includes collaborations for developing digital skills and attracting investments in technology.

For Ukraine, strengthening digital infrastructure represents an important component of the reconstruction process, having the potential to support the development of public services, the digital economy, and international investments.

The partnership confirms the increasingly important role of software and telecom infrastructure in national economic development and resilience strategies.

The European Union intensifies its investigation into Meta's practices

The European Commission continues its investigation into Meta regarding how its platforms may encourage behaviors considered addictive among users, especially minors. The analysis is part of the implementation of the Digital Services Act.

European authorities are examining elements such as content recommendation mechanisms, notifications, reward systems, and other features that may influence the time spent on platforms. The investigation aims to establish whether Meta complies with user protection obligations and assesses systemic risks.

For digital companies, the investigation represents a new example of how European regulations are beginning to influence the design of software products. Aspects such as interface design, recommendation algorithms, and user experience are becoming increasingly relevant from a legislative compliance perspective.

The outcome of the investigation could have implications for the entire industry, as the standards established in the European Union are often subsequently adopted in other markets.

ChatGPT understands user intent better and responds more naturally

OpenAI has introduced significant improvements in how ChatGPT interprets user requests. The update allows the model to better understand the context of a conversation and the intent behind a request, reducing situations where responses follow instructions in an excessively literal manner.

The new approach enables the model to fill in missing information when the context is clear enough and to provide responses that are closer to how people communicate. At the same time, ChatGPT can better identify situations where clarifications are needed before generating a response, avoiding misinterpretations of ambiguous requests.

For users and companies, this evolution means more natural interactions and a reduction in the time spent reformulating prompts. As AI models are integrated into business applications, virtual assistants, or software products, the ability to understand user intent becomes as important as the technical performance of the model.

The update reflects the direction in which conversational models are evolving. The focus is gradually shifting from strictly executing instructions to understanding context, collaborating with the user, and providing more relevant responses in real situations.

MacBook with touchscreen seems closer to launch

New information from the supply chain indicates that Apple is continuing to develop its first touchscreen MacBook, a product that has been discussed for many years. According to reports, the future device is expected to include two significant modifications: an OLED screen and full support for touch interaction.

Until now, Apple has consistently maintained that the user experience on Mac is optimized for keyboard and trackpad, while touch functions are reserved for the iPad. However, the evolution of hardware and the proximity between iPadOS and macOS suggest that this strategy may begin to change.

A MacBook with a touchscreen could open new scenarios for professional applications, creative tools, and multitasking, without fundamentally altering the identity of the operating system. For developers, the emergence of a new method of interaction could also lead to the adaptation of existing interfaces and workflows.

Although Apple has not officially confirmed the existence of the product, recent information indicates that the project continues to advance and that the premium laptop market could enter a new stage of competition.

China leads the global supercomputer ranking

China ranks first in the world in terms of the number of high-performance supercomputers, according to a recent analysis dedicated to high-performance computing infrastructure. The country has consolidated its investments in recent years, developing both its own processors and computing systems capable of supporting scientific research and the development of artificial intelligence.

Supercomputers are used for complex simulations, medical research, the development of new materials, climate forecasting, and training large AI models. In the context of global competition for digital infrastructure, processing capacity becomes a strategic advantage comparable to access to energy resources or semiconductors.

China's investments reflect the growing importance of computing infrastructure in the digital economy. As AI models grow in size and complexity, access to sufficient computing power becomes one of the main factors influencing the pace of innovation.

The competition for the development of supercomputers continues to be closely linked to national strategies regarding artificial intelligence, semiconductors, and technological independence.

Mayors of major cities call for stricter rules for AI data centers

Over 40 mayors from around the world are calling for stricter measures regarding the development of data centers dedicated to artificial intelligence, citing their impact on energy consumption, water resources, and urban infrastructure.

Local administration representatives argue that the accelerated pace of data center construction generates significant pressures on electrical networks and available resources in local communities. They are calling for more careful planning and greater transparency regarding the resource consumption of these facilities.

As AI becomes increasingly dependent on computing infrastructure, data centers are transformed into a strategic element of the digital economy. At the same time, their expansion raises questions about sustainability and how technological investments can be balanced with the needs of communities.

The initiative reflects a global trend where the development of AI infrastructure is beginning to be analyzed not only from the perspective of technological performance but also of environmental impact and urban development.

UN calls on AI companies to disclose environmental impact

The Secretary-General of the United Nations is calling on companies developing artificial intelligence to be more transparent regarding energy consumption, carbon emissions, and the use of resources necessary for training and operating AI models.

According to the UN, the rapid development of artificial intelligence brings significant benefits, but the environmental costs associated with data centers and the necessary infrastructure remain insufficiently documented. The organization believes that reporting this information should become a standard practice in the industry.

For companies, the request highlights the importance of sustainability in the development of digital infrastructure. Energy efficiency, the use of renewable sources, and optimizing resource consumption are becoming increasingly important criteria for both investors and regulatory authorities.

As AI is adopted on a large scale, discussions about performance are complemented by concerns about environmental impact, and transparency becomes an essential element of responsible development.

Samsung introduces fees for using SmartThings APIs

Samsung has announced that it will begin charging for access to certain SmartThings APIs used by developers and commercial partners. The change aims to support the development of the platform and reflects the rising costs associated with operating cloud services and IoT ecosystems.

SmartThings APIs allow for the integration of smart devices with third-party applications and represent one of the important components of the ecosystem developed by Samsung for smart homes. The new fees will primarily target commercial use and large volumes of requests.

For software developers and companies building products integrated with SmartThings, the change may influence operating costs and business models. At the same time, the change reflects the increasingly common trend of platforms monetizing the technical infrastructure provided to partners.

The decision confirms the maturation of IoT ecosystems, where the development and maintenance of the infrastructure necessary to connect millions of devices require significant investments and sustainable business models.

Canva expands its AI platform to the advertising industry

Canva continues to expand its portfolio of artificial intelligence-based products and announces new tools dedicated to the advertising industry. The company aims to transform the platform from a design solution into a complete ecosystem for creating, managing, and distributing digital campaigns.

The new functionalities allow for the rapid generation of advertising materials tailored to different platforms, automatic resizing of creatives, and optimization of content for multiple formats. AI tools are integrated directly into the workflow, reducing the time required for campaign production and facilitating collaboration between marketing and creative teams.

For companies and agencies, Canva's expansion reflects the direction in which software applications dedicated to digital content are evolving. Artificial intelligence is becoming a standard component of the creative process, automating repetitive activities and allowing teams to focus more on strategy and concept.

Through this initiative, Canva enters direct competition with established platforms in the marketing and professional design space, confirming that AI is rapidly redefining how digital campaigns are produced and managed.

Nearly 400 local newspapers sue OpenAI and Microsoft

A group of nearly 400 local publications in the United States has filed a lawsuit against OpenAI and Microsoft, claiming that artificial intelligence models have been trained using journalistic content without the consent of the publishers.

The plaintiffs argue that the use of articles in the training process of AI models affects the value of original content and can reduce traffic to publications that invest in producing information. The lawsuit adds to other similar actions initiated in the past two years by media organizations and copyright holders.

For the AI industry, such litigation can influence how future models will be built and the commercial relationships between technology developers and content creators. More and more companies are already choosing to enter into licensing agreements with publishers and media organizations for legal access to data.

The case reflects one of the most important legal debates generated by artificial intelligence: how to balance technological innovation with the protection of copyright and the sustainability of the media industry.

Chrome modernizes the Autofill feature with the help of artificial intelligence

Google is preparing a significant update for the Autofill feature in Chrome, which will use artificial intelligence for more accurate and contextual completion of online forms.

The new version will not be limited to identifying standard fields, such as names or addresses, but will be able to better interpret the structure of pages and the type of information requested. The goal is to reduce errors and simplify the form-filling process, regardless of how they are constructed.

For web application developers, the update underscores the importance of adhering to modern standards regarding forms and accessibility. At the same time, users will benefit from a faster and more consistent experience in interacting with online services.

The integration of AI into seemingly simple functions like Autofill demonstrates that artificial intelligence is beginning to be present in almost all components of software products, including those used daily by millions of people.

CTOs' confidence in the ability to scale AI continues to decline

Confidence among technology directors in organizations' ability to implement and scale artificial intelligence has decreased for the third consecutive year, according to the latest report by Akkodis. Although investments in AI continue to grow, more and more companies are facing difficulties in transforming pilot projects into widely used solutions.

The report shows that the main challenges are integration with existing systems, data quality, governance, security, and the lack of necessary skills for managing complex implementations. Access to high-performance AI models is no longer the main obstacle, but rather their integration into real business processes.

For organizations, the findings confirm that the success of AI implementation depends on infrastructure, software architecture, and change management, not just on choosing a high-performing model. Companies that invest in governance and modernizing existing systems are better prepared to achieve long-term results.

The report reflects the maturation of the AI market. After a period of experimentation, attention is shifting towards scalability, reliability, and the integration of technology into critical operations of organizations.

AWS: The best way to learn AI is through experimentation

Julia White, Chief Marketing Officer at Amazon Web Services, believes that organizations and professionals who want to leverage artificial intelligence must constantly experiment and accept that initial attempts will not always yield the desired results.

At the VivaTech conference, she explained that the first AI implementations brought moderate improvements in productivity, but the truly significant benefits only appeared after a complete rethinking of workflows. Simply adding AI on top of existing processes produces limited results, while redesigning the way of working can generate significantly greater gains.

The message is relevant for companies implementing AI in software development, operations, or administrative activities. The value of technology emerges when organizations are willing to adapt processes and experiment with new ways of working, not just automate existing steps.

The statements confirm one of the current trends in the market: AI adoption is less of a technological challenge and increasingly an exercise in organizational transformation.

Artificial intelligence deciphers a nearly 2,000-year-old scroll

An international team of researchers has managed to read approximately 20 columns from a carbonized scroll recovered from the Roman city of Herculaneum, using a combination of X-ray imaging and artificial intelligence algorithms. The document was sealed by the eruption of Vesuvius in 79 AD and cannot be unrolled without being destroyed.

The technology used identifies very fine differences between papyrus fibers and invisible ink traces, allowing for the digital reconstruction of the text without damaging the artifact. The first deciphered fragments contain philosophical texts associated with the Stoic current and provide new information about ancient thinking.

The result demonstrates how artificial intelligence can accelerate scientific research in very different fields from software or business. Image analysis algorithms are becoming valuable tools for archaeology, history, and cultural heritage preservation.

The Vesuvius Challenge project continues to analyze other carbonized manuscripts, and researchers hope that the same technology will allow, in the coming years, the recovery of content from hundreds of documents previously considered impossible to read.

Europol warns of the rise of technology-supported organized crime

Europol has published a new report on the most dangerous criminal networks operating in the European Union, highlighting the increasingly important role of technology and artificial intelligence in the activities of organized crime groups.

The report identifies 731 active networks, with over 400,000 members from 118 nationalities. Drug trafficking remains the main criminal activity, but online fraud, cyberattacks, and the exploitation of digital infrastructures are growing at an accelerated pace. Europol emphasizes that criminal organizations are using AI, digital tools, and global financial systems to expand their operations and avoid detection.

For companies and public institutions, the report's conclusions highlight the importance of investing in cybersecurity, monitoring digital infrastructures, and strengthening fraud detection mechanisms. Digitalization brings significant benefits, but at the same time expands the attack surface available to malicious actors.

The report confirms that the fight against organized crime is becoming increasingly dependent on international cooperation and the use of advanced technologies to identify and prevent threats.

Intensive use of AI remains low in eurozone companies

Although over 70% of companies in the eurozone report that they use artificial intelligence or intend to start implementing it during the year, only 7% use AI intensively, according to a study conducted by researchers at the European Central Bank.

The analysis, based on responses from over 5,000 companies, shows that advanced use of AI is particularly found in small, young, and service-oriented organizations. In contrast, large companies adopt the technology at a slower pace, and in many cases, usage remains limited to isolated projects or experimental activities.

The results suggest that there is still significant potential for expanding the use of AI in the European economy. Although interest in technology is high, transforming business processes and integrating AI into essential operations requires investments in infrastructure, skills, and governance.

The study confirms that the gap between initial adoption and intensive use remains one of the most important challenges for European organizations, and the economic benefits on a large scale will depend on companies' ability to integrate AI into their daily activities.

Synthesis made with the help of a monitoring flow provided by Control F5 Software.