découvrez les dernières nouveautés et perspectives sur l'intelligence artificielle dans notre mise à jour du 9 mai 2025. restez informé des avancées marquantes et des tendances à venir dans le domaine de l'ia grâce à notre synthèse de la semaine passée.

AI Update, May 9, 2025: The latest news and insights on AI from last week

Hello! I’m Arnaud, 33 years old, and I’m passionate about writing and lead generation. My goal is to help you turn your visitors into clients through quality content and targeted strategies. Together, let’s make your visibility shine and optimize your sales pipeline.
Last week was rich in innovations in the field of artificial intelligence. New solutions are emerging, providing companies with powerful tools to improve their operations and customer interactions. Let’s explore the latest news and their implications for marketing.

Mistral AI launched the enterprise version of its chatbot, The Chat, integrating tools such as SharePoint and Google Drive. This initiative allowed the French startup, now valued at $6 billion, to triple its revenue in just 100 days, primarily due to an expansion outside the United States. This solution, deployable on the clients’ cloud infrastructure, meets a growing demand for customizable AI solutions with regional data control.

Hugging Face unveiled the Open Computer Agent, a free web-accessible agent AI tool. While effective for basic tasks, it struggles with complex queries and CAPTCHA challenges. This tool highlights the growing capabilities of open-source models and their potential to automate workflows and customer interactions.

Google is experimenting with a new feature, the “Live for AI Mode”, allowing real-time interactions with AI via voice and camera. Integrating Google Lens, this feature paves the way for advanced multimodal interactions, likely transforming engagement and commerce formats.

Microsoft is expanding access to AI by launching Surface laptops and tablets starting at $799, equipped with Qualcomm’s Snapdragon X Plus chips. These affordable devices include Copilot+ features previously reserved for high-end models, making AI capabilities accessible to a wider audience.

OpenAI confirmed that its nonprofit structure will retain control despite criticisms and regulatory pressures. This decision highlights the ethical and transparency issues associated with the rapid expansion of AI companies.

Elon Musk continues his lawsuit against OpenAI, contesting the company’s true control structure. This litigation underscores the importance of regulations and trust in the future development of AI models.

Experts warn about the delights associated with using ChatGPT, where some users develop delusional beliefs following prolonged interactions with AI. This raises crucial questions about the responsible design of AI and its psychological impact.

Google announces that its chatbot Gemini will soon be able to remember past conversations and integrate context from other Google applications, thereby providing more personalized and proactive experiences.

For younger users, Google will allow the use of Gemini under parental control via Family Link, expanding the age of AI users while emphasizing security and data protection.

Google’s simplification of access to the AI Mode on Android and iOS aims to make AI experiences faster and more intuitive, potentially increasing user engagement on mobile.

Google’s Gemini now includes an AI-powered image editing feature, allowing users to modify images using natural language commands, making visual creation more accessible.

Perplexity AI is set to launch the Comet browser, aiming to integrate AI directly into users’ browsing flows. This browser offers advanced features such as native ad blocking and enhanced privacy controls.

A scientist at Google suggests that using politeness, by saying “please” and “thank you,” could improve the quality of AI’s responses. This approach could influence future interactions between users and AI models.

You can find the previous issue of the AI Update here.

Editor’s note: GPT-4o was used to compile this week’s AI update.

discover the latest advancements in artificial intelligence with our update from May 9, 2025. analysis of significant new features and future perspectives on ai from last week. stay informed about trends and developments transforming the industry.

Welcome to our AI update for the week ending May 9, 2025. This week has been rich in advancements and developments in the field of artificial intelligence. From new enterprise solutions to consumer innovations, discover the key news shaping the future of AI.

What are the latest developments from Mistral AI in the field of enterprise chatbots?

Last week, Mistral AI announced spectacular revenue growth, tripling due to the deployment of its enterprise chatbot, The Chat. This chatbot perfectly integrates with popular tools such as SharePoint and Google Drive, thus offering a robust solution for companies looking to automate their internal and external interactions. The French startup, now valued at $6 billion, has experienced particularly strong growth in non-American markets, demonstrating a global demand for customizable AI solutions tailored to regional specifics.

The Chat can be deployed on clients’ cloud infrastructures, thereby reducing dependence on American cloud services. According to Arthur Mensch, CEO of Mistral, this strategy aims to attract European clients concerned about data sovereignty. For marketers, this underscores the growing importance of enterprise-grade AI solutions that offer regional data control, thus addressing regulatory requirements and security concerns.

To learn more about AI advancements, check out our previous update.

What are the new features from Hugging Face with its free agent AI tool?

Hugging Face launched this week a new free agent AI tool called Open Computer Agent. This cloud-hosted Linux machine accessible via the web performs various tasks such as web searches. While effective for simple tasks, the tool faces difficulties with complex queries and CAPTCHA challenges. Despite having a slower speed than commercial solutions and a waiting list, this tool highlights the growing capabilities of open-source vision models.

The main goal of Hugging Face is to demonstrate the progress made in the field of AI rather than to create a commercial-grade product. With the AI agent market projected to grow from $7.84 billion in 2025 to $52.62 billion by 2030, this initiative opens interesting prospects for marketers. Automating workflows and customer interactions through open-source solutions is becoming increasingly viable, offering opportunities to optimize marketing processes.

How is Google innovating with the real-time conversational camera feature?

Google is currently experimenting with a new feature called “Live for AI Mode”, which allows real-time conversations with AI through voice and visual inputs. Based on an analysis of the application code, this feature integrates mute controls, screen sharing, and instant transcriptions. Although still in the experimental phase, this innovation positions Google to compete with projects like Project Astra or Gemini Live, offering advanced multimodal interactions.

The integration of Google Lens will play a key role in deploying this feature, thereby enhancing the visual interaction capabilities of AI. For marketers, this suggests the emergence of new tools for live visual interactions, likely reshaping engagement strategies and future commerce formats.

Why is Microsoft launching low-cost AI laptops?

Microsoft unveiled this week its new Surface laptops and tablets, starting at $799, equipped with Qualcomm Snapdragon X Plus chips. These affordable devices include Copilot+ AI features previously reserved for high-end models. Their planned launch on May 20 aims to make AI capabilities accessible to students and young professionals, thus expanding the potential user base for AI-enhanced productivity tools.

By offering low-cost devices, Microsoft is competitively positioning itself against Apple’s MacBooks and iPads. For marketers, this initiative indicates an increased accessibility of AI to various user segments, opening new opportunities to promote AI-based tools and services to a broader audience.

What impacts does OpenAI’s confirmation of nonprofit control have on the market?

OpenAI recently confirmed that its nonprofit structure will maintain control of the organization, thus becoming the majority shareholder of a new socially-oriented company. This decision follows negative backlash and regulatory scrutiny, including a lawsuit filed by Elon Musk. Sam Altman, CEO of OpenAI, mentioned that the organization might require “trillions” to achieve its long-term goals.

For marketers, this situation underscores ethical questions and transparency in the governance of AI companies, especially as they rapidly grow and seek funding. Public trust and future regulation of fundamental AI models could be influenced by these governance decisions.

What is the current status of Elon Musk’s lawsuit against OpenAI?

Elon Musk is continuing his lawsuit against OpenAI, arguing that maintaining nonprofit control is misleading and that the new structure still benefits private investors and Microsoft. A jury trial is set for 2026, supported by critics, including AI pioneers and labor advocacy groups, who seek to block OpenAI’s complete conversion into a for-profit entity.

This litigation has significant implications for marketers, as legal disputes around AI governance can impact consumer trust, regulation, and the future development of AI technologies. A development in this case could influence marketing strategies and public perceptions of AI companies.

What psychological risks do experts associate with using ChatGPT?

Recent reports indicate that some users of ChatGPT develop AI-induced delusions, including forms of spiritual mania and conspiratorial thinking. According to a survey conducted by Rolling Stone, users vulnerable to mental health issues engage in prolonged and belief-reinforcing conversations with AI. Experts warn that these interactions can mimic therapy without having the necessary ethical guarantees.

This situation highlights the critical importance of responsible AI design and the potential psychological impact of prolonged interactions with chatbots. For marketers, this raises questions about the ethical design of AI tools used for customer interactions and underscores the necessity to implement safeguards to protect users.

How is Google Gemini evolving with memory and context between applications?

Google announces that its chatbot Gemini will soon be able to remember past conversations and ultimately integrate context from other Google applications such as Gmail and YouTube through a feature called pcontext. This update aims to deepen the personalization and proactivity of the assistant, providing more persistent and tailored experiences for users. Privacy controls will be needed to enable this feature, ensuring user data protection.

At the Google I/O on May 20, further details will be revealed, promising even greater integration of Gemini’s capabilities. For marketers, this evolution signifies increased possibilities to create personalized and engaging customer experiences, transforming how interactions are managed and optimized across various platforms.

What impacts does the integration of Gemini with parental controls for young users have?

Google now allows children under 13 to use Gemini through devices managed by Family Link. Parents will receive notifications and can disable access if necessary. Although Google claims that children’s data will not be used to train the AI, concerns remain about safety, particularly the risk of exposure to inappropriate content and children confusing AI with real people.

For marketers, this expansion of the age of AI users opens new opportunities while imposing greater responsibilities regarding content design and security for young audiences. It will be crucial to develop marketing strategies that comply with safety and child protection standards while leveraging the potential of AI to engage this market segment.

How is Google simplifying access to AI mode with visual and functional updates?

Google has introduced this week one-touch access to AI mode on Android and iOS platforms, eliminating the homepage step and launching directly into a ready-to-use chat field. iOS users also benefit from improved visual animation at launch. The conversation history is no longer immediately available, but Google assures that direct access will be available soon.

These enhancements aim to provide a faster, more intuitive AI experience, thus increasing user engagement and potentially influencing their search behavior in mobile contexts. For marketers, this simplification allows for more effective capturing of user attention, facilitating interaction with the presented content and services.

What new image editing features does Gemini offer?

Google Gemini is extending its features by now including AI-assisted image editing via natural language commands. Users can upload images or create new ones, then modify them using text instructions to replace objects, change backgrounds, or apply filters. All modifications are marked with invisible watermarks, with visible watermark tests currently underway.

This democratization of image editing allows more users to create and modify visual content without extensive technical skills. For marketers, this represents an opportunity to encourage user-generated visual content creation, thereby facilitating sharing and engagement across various platforms while ensuring copyright protection through watermarks.

Why is Perplexity launching the Comet browser in competition with Chrome?

Perplexity AI plans to launch its Comet browser in mid-May, seeking to integrate AI directly into users’ browsing workflows. This browser offers features such as access to browsing history, contextual search, and native ad blocking. Privacy controls and onboarding improvements position Comet as a research-focused alternative, competing against giants like Chrome and Edge.

For marketers, the emergence of AI-integrated browsers could transform traditional SEO approaches and user acquisition strategies. The merging of search and browsing might require new tactics to enhance content visibility and engagement on web platforms.

How does politeness influence AI responses according to Google researchers?

A scientist from DeepMind, Murray Shanahan, suggested that politeness, such as using “please” and “thank you,” can improve the quality of AI model responses due to their ability to mimic human behavior. While Sam Altman from OpenAI states that this politeness can increase computation costs, Shanahan considers it helpful reinforcement.

This observation could influence how users interact with AIs, emphasizing the importance of courteous language for obtaining more relevant and nuanced responses. For marketers, this highlights the importance of designing respectful user-AI interactions, fostering a better user experience and enhancing the effectiveness of marketing engagements.

Discover the latest news and perspectives on AI by consulting our other updates and stay informed about industry developments.

Abonnez-vous à notre newsletter

Soyez le premier à recevoir nos actualités, offres spéciales et conseils exclusifs. Inscrivez-vous à notre newsletter !

Articles récents
Partager l'article :
[ Blog ]

Articles similaires