Update on AI from June 27, 2025: the news and perspectives of the past week

Advancements in artificial intelligence continue to redefine our technological landscape. The past week was rich in court decisions influencing the development of AI. Giants like Meta and OpenAI are achieving significant legal victories.
These developments open new perspectives for experts in marketing, ensuring a smoother integration of AI tools. With the recent updates from platforms such as Salesforce and Creatio, lead generation strategies are becoming more effective. However, concerns regarding data security, such as those revealed by Scale AI, highlight the importance of information governance. This week proves crucial for anyone wishing to stay at the forefront of innovations in digital marketing.

discover the latest news and exciting perspectives on artificial intelligence in our update from June 27, 2025. stay informed about key advancements of the week and explore the issues shaping the future of technology.

Table des matières

Anthropic wins a major court ruling on fair use in AI training

This week, Anthropic achieved a significant victory in a court case regarding the fair use of literary works for training artificial intelligence models. A U.S. federal judge ruled that Anthropic’s use of books to train its Claude model fell under the concept of « fair use, » thereby establishing a crucial legal precedent for AI companies.

The court emphasized that the use of books was « extremely transformative, » although storing copyrighted works in a central library was not permitted. This ruling could influence ongoing lawsuits against giants like OpenAI, Meta, and others, potentially alleviating copyright concerns in AI development.

For marketers, this decision is of utmost importance. If affirmed, it would allow AI providers to continue training their models on publicly available content, thereby preserving the pace of innovation that fuels marketing tools. For more information about recent developments in AI, check out the April 2025 update.

Meta and OpenAI win rulings on fair use in copyright lawsuits

Meta and OpenAI both secured favorable rulings in copyright-related lawsuits. Meta triumphed in a lawsuit brought by 13 authors, including Sarah Silverman, with the judge finding that their arguments did not demonstrate economic harm. This decision ends the case without validating all AI training practices but establishes a precedent where creators cannot prove direct economic damages.

This ruling echoes that of Anthropic obtained earlier in the week, further strengthening the position of AI companies against copyright claims. For marketing professionals, these multiple victories reduce short-term legal risks associated with AI tools, encouraging a more aggressive integration of products and content capabilities. This trend is also addressed in the March 2025 update.

Salesforce enhances its Agentforce platform with global model options and increased observability

Salesforce unveiled version 3 of its Agentforce platform, integrating advanced features such as real-time monitoring, multi-agent coordination, and support for the Model Context Protocol (MCP). These improvements allow for secure integration with external tools, expanding options for regulated industries thanks to support for Anthropic and Google Gemini models.

Among the early adopters is PepsiCo, which is already benefiting from the platform’s new capabilities. For marketers, these enhancements facilitate the enterprise-level deployment of AI in marketing operations and customer experience (CX) management, ensuring consistent agent performance and rigorous data governance. For more details, refer to the February 2025 update.

Scale AI faces security concerns after leaking sensitive data

Scale AI has recently been pointed out for major security lapses. Business Insider revealed that Scale AI regularly used public Google Docs to manage projects with clients like Meta, Google, and xAI, thus exposing thousands of contractor emails, payment disputes, and even confidential training documents to anyone with the link.

This exposure highlights the critical importance of data governance when collaborating with AI partners. Brands using AI training or annotation services must closely examine their providers’ security protocols to avoid such vulnerabilities. More information is available in the December 2024 update.

Gartner forecasts cancellation of 40% of agentic AI projects by 2027

According to a recent study by Gartner, nearly 40% of agentic AI projects will be canceled by 2027 due to rising costs and uncertain return on investment (ROI). Many current initiatives remain experimental, and numerous vendors practice what Gartner calls « agent washing, » dressing up existing tools with agentic claims without offering true capabilities.

For marketers, this forecast emphasizes the necessity of carefully scrutinizing the promises of agentic AI tools and investing only where these tools clearly enhance workflows or outcomes. Prudently managing AI projects is essential to avoid costly failures and maximize potential benefits. Discover more perspectives in the January 2025 update.

Creatio launches a native CRM update for AI, challenging Salesforce’s AI add-ons

Creatio announced version 8.3 Twin of its CRM, integrating AI natively. This update introduces conversational interfaces, no-code agent building, and persistent multi-channel interactions. Priced at $25 per user per month, the platform offers AI features without requiring upsells, making AI more accessible to medium-sized teams.

For marketers, this evolution reflects a growing trend towards integrated and affordable AI solutions within marketing platforms, reducing complexity and fostering adoption among intermediate segment teams. For an in-depth analysis, consult the December 2024 update.

Perplexity reaches a $14 billion valuation despite criticism and interest from tech giants

Perplexity AI recently raised $500 million, reaching a valuation of $14 billion. Despite critiques labeling it a glorified API aggregator, Perplexity’s sleek interface and ad-free search engine attract many users. Tech giants like Apple and Meta have expressed interest in a potential acquisition.

For marketers, widespread adoption of Perplexity could disrupt the way consumers and professionals discover content, leading to changes in SEO strategies and content visibility. More details on this dynamic can be found in the April 2025 update.

OpenAI considers adding ads to ChatGPT to offset operational costs

OpenAI‘s CEO, Sam Altman, announced that ads could be introduced in future versions of ChatGPT to cover rising infrastructure costs. He emphasized a cautious approach to ad deployment, ensuring not to compromise user trust.

This development represents a new opportunity for marketers to integrate advertising placements within AI-generated content while presenting challenges in brand message control and visibility. For a complete understanding of the implications, consult the March 2025 update.

Elon Musk plans to « rewrite human knowledge » by retraining Grok

Elon Musk announced his intention to rewrite human knowledge by retraining his Grok model with a revised corpus, eliminating what he qualifies as « waste » in the current data. This initiative aims to refine the quality of information provided by Grok, though critics highlight the risk of historical distortion akin to « rewriting history. »

For marketers, if Grok becomes a widely used AI platform, the unique training data could influence access to information, particularly on controversial or politicized topics. This may necessitate adjustments to content strategies and a reevaluation of data sources used.

Samsung considers subscription levels for Galaxy AI features in 2026 rollout

Samsung plans to introduce a paid subscription model for its Galaxy AI features starting in 2026. While not all features will require payment, some of the most popular will remain free. This approach allows Samsung to monetize its AI innovations while providing added value without excluding non-paying users.

For marketers, this model signals a broader trend towards monetizing AI features in consumer technology. It affects how reach strategies, access, and platform capabilities are envisioned in digital marketing. More details can be found in the February 2025 update.

OpenAI warns of Zhipu AI’s advancement in China in the international market

OpenAI has alerted about the advances of Zhipu AI in China, which is making international contracts with governments and companies. Backed by $1.4 billion in state funding, this startup is part of a broader effort to assert Chinese influence via sovereign AI.

For marketers, the rise of Chinese alternatives in AI could impact access to the global market, partnerships, and regulatory alignment, especially for international campaigns and decisions regarding technology stacks. To explore these implications, refer to the April 2025 update.

The U.S. Senate debates a moratorium on AI regulation tied to broadband funding

A controversial bill in the U.S. Senate proposes to block states from regulating AI in exchange for funding for broadband infrastructure. Critics, including members of the Teamsters union, denounce this initiative as a « gift » to large tech companies, while supporters argue that it avoids a patchwork of state laws.

If this bill were to pass, it could accelerate national AI deployments by reducing compliance complexity, but it could also fuel consumer distrust in lightly regulated environments. marketers will need to navigate this evolving landscape, balancing innovation and customer trust. For more information, consult the January 2025 update.

Sam Altman criticizes the NYT lawsuit and defends ChatGPT’s private data management

In a live interview on a podcast, Sam Altman, CEO of OpenAI, strongly criticized the lawsuit filed by The New York Times seeking the preservation of user logs in the copyright case. Altman labeled this demand as an attack on user privacy.

This ongoing legal situation and changing public narratives may influence brands’ decisions regarding the use or support of OpenAI tools in their content pipelines. marketers must stay informed about legal developments to adjust their strategies and ensure compliance with privacy standards. For a detailed analysis, refer to the March 2025 update.

Meta adds AI summaries to WhatsApp, powered by private processing

WhatsApp is rolling out a new optional feature for message summaries powered by AI, beginning in the United States. This feature utilizes Meta’s AI while preserving user privacy through « Private Processing. »

For marketers, this innovation opens the door to smarter customer service interactions on WhatsApp, a key channel in global conversational marketing. Businesses can now provide faster and more personalized responses, thus enhancing customer experience and engagement. For more details, consult the April 2025 update.

OhChat launches AI celebrity-powered digital twins for paid interactions

OhChat now offers customized AI avatars of celebrities, providing paid interactions that can sometimes be sexualized. The platform already boasts 200,000 users and raises concerns regarding emotional dependency and the ethics of digital work.

For marketers, these digital twins represent a new form of monetization of influencer engagement, with the potential to be soon applied to brand ambassadors or corporate spokespersons. This trend opens new avenues for immersive and interactive marketing campaigns. More information is available in the January 2025 update.

Apple considers acquiring Perplexity AI to enhance its search capabilities

Reports indicate that Apple is exploring the acquisition of Perplexity AI in order to reduce its dependence on Google for search services. This strategy aligns with Apple’s goal to further integrate AI within its ecosystem, offering more personalized and efficient user experiences.

For marketers, if this acquisition comes to fruition, it could create a third major content discovery platform to optimize alongside Google and OpenAI. The emergence of such a platform would influence SEO strategies and the visibility of digital content. For more context, consult the April 2025 update.

Perplexity expands access to its AI-powered Comet browser

The Comet browser from Perplexity, centered on AI, is now available for some Windows users and may soon come to Android. Its features include virtual clothing try-ons, discount checks, and agentic searching.

For marketers, these new browsing and shopping experiences may lead to changes in advertising targeting, product visibility, and customer expectations. Rapid adaptation to these technological evolutions will be essential to maintain a competitive presence. For more information, consult the January 2025 update.

Getty drops copyright claims against Stability AI in a UK case

Getty Images has withdrawn its main copyright claims in its lawsuit against Stability AI in the UK. This decision highlights the difficulty of proving infringement when AI training occurs outside national borders.

For marketers, this development could encourage more AI companies to train their models on vast corpuses of content, including media, without requiring licenses. This presents both risks and opportunities for marketing, especially concerning content creation and copyright management. More details are available in the January 2025 update.

A MIT study suggests using AI may lead to « cognitive debt »

A small study conducted by the MIT revealed that users writing essays with ChatGPT showed lower brain engagement compared to control groups. However, critics of the study point to the limited experimental design, rendering the conclusions shaky.

For marketers, this research emphasizes the importance of finding a balance between AI assistance and human creativity in order to maintain engagement and learning within teams. Thoughtful use of AI can maximize its benefits while minimizing potential negative impacts. For more information, consult the December 2024 update.

YouTube rolls out AI-powered search previews for Premium users

YouTube is currently testing AI-generated summaries in search results for categories such as travel and shopping. This feature is currently available on mobile for English-speaking Premium users.

For marketers, this feature could reduce clicks to organic content, thereby prompting the need to optimize video titles and metadata for AI summarization. Strategic adaptation will be necessary to maintain visibility and engagement for videos in this new environment. For more information, consult the December 2024 update.

LinkedIn observes a sixfold increase in AI-related jobs and a twentyfold increase in AI skills

LinkedIn has seen an exponential increase in job postings mentioning AI, with a sixfold increase in positions and a twentyfold increase in AI skills within members’ profiles. CEO Ryan Roslansky noted that the flow of activity on the platform now functions « like a large prompt. »

For marketers, this intense demand for AI talent encourages a reevaluation of recruitment, training, and skill development strategies within marketing departments. Investing in AI skill development can offer a significant competitive advantage. For more details, consult the April 2025 update.

Pangram raises $4 million to expand detection of AI-generated texts

The startup Pangram, specialized in detecting AI-generated texts, has raised $4 million to expand its team and offerings. Founded by former engineers from Tesla and Google, Pangram serves clients such as Quora and NewsGuard.

With the proliferation of AI-generated content, detection tools like Pangram are becoming essential in sensitive sectors like education, media, and reputation management. marketers will need to integrate these tools to ensure the authenticity and quality of broadcast content. For more information, consult the April 2025 update.

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

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

découvrez notre partenariat officiel, garantissant qualité et confiance pour des collaborations durables et performantes.

Devenez Partenaire Officiel

Le rendez-vous incontournable des leader·e·s féminines se profile à l’horizon. Votre marque peut prendre une place de choix dans cette