It may sound like a joke, but it is not. Pexplexity published an official statement entitled “Bullying is not innovation”, in which it accuses Amazon of intimidating it so that users of Comet stop using the tool as an intelligent shopping assistant on the marketplace.

According to the startup, the objective of the online retail giant is to eliminate users’ rights in order to sell more advertisements now and to partner with AI agents designed to exploit users in the future. “It is not just harassment: it is madness,” the document asserts.

In response, Andy Jassy, CEO of Amazon, stated that Perplexity should respect its wish for the shopping agent to avoid its website, in order to “ensure a positive customer experience.” According to the executive, external AI agents lack personalization and purchase history, and therefore provide inaccurate prices and incorrect delivery times.

Who holds the power: users or the Marketplace?

Here, the discussion goes beyond a mere corporate dispute: can you use your own AI assistant to shop on Amazon, or only those that the marketplace approves? According to Perplexity, the answer Amazon is delivering is a clear “no.”

Let us recall that Amazon has its own AI-based shopping assistant, Rufus, which we have recently discussed, and which has not gained much popularity worldwide. In fact, the country most suspicious of the tool is the United States itself, birthplace of Amazon. Despite having been launched less than two years ago, the marketplace has not succeeded in securing natural and frequent use of Rufus among its users. The arrival of Comet further complicates its plans to strengthen its native assistant.

In this regard, Perplexity asserts that, although websites have always had the right to block bots attempting to crawl them, it is not unusual for major technology companies to try to impede potentially disruptive tactics employed by new competitors. “Amazon generates tens of billions annually by selling advertising space on its marketplace, a highly profitable source of revenue that would decrease if people began to use agents to shop instead of browsing themselves,” it argues.

The influence of an AI Assistant on purchase intentions

Perplexity poses the same question: “The rise of agentive AI presents a choice: will this technology empower users to control their digital lives, or will it become another tool for corporations to manipulate and exploit them?”.

The startup maintains that tools such as Comet enable users to save time, make better decisions, and simplify the purchasing process, since these assistants are not distracted by advertisements nor are they swayed by sponsored products. They go straight to the point.

Amazon justifies its decision with reasons related to quality: it claims that external assistants provide unreliable data, incorrect delivery estimates, and prices that might be outdated.

Perplexity denies this and views it as a means of stifling innovation and maintaining complete control over how consumers navigate and make purchases within its platform: “The Comet Assistant quickly finds and buys the item for you, saving you time for more important tasks (…) But Amazon does not care. It is more interested in showing you advertisements, sponsored results, and influencing your purchasing decisions with upsells and confusing deals.”

How AI Agents should be

According to Perplexity’s perspective, personal agents represent the future of online navigation. “They are distinct from crawlers, collectors, or bots. A user agent is your AI assistant: it has exactly the same permissions as you, acts only when you request it, and solely on your behalf,” it explains, and adds that in order for user agents to fulfill their true purpose, they must be private, personal, and powerful in order to be genuinely useful for users.

Therefore, in response to Amazon’s demand, Perplexity asserts that it will not be intimidated and will continue to fight for users’ rights. “People love our products because they are designed for them. User freedom and choice are at the heart of everything we build,” it states, while addressing a pointed comment at Jeff Bezos: “Amazon also forgets how it became so large: users love it. They want good products at a low price and delivered quickly. Agentive shopping is the natural evolution of that promise, and people are already demanding it. Perplexity demands the right to offer it.”

Photo: Gemini

The post Perplexity accuses Amazon of “bullying” for attempting to prohibit the use of Comet in the marketplace appeared first on Marketing4eCommerce English.

Had you heard of Rufus before reading this article? If your answer was “no,” then you are among the majority (and we had already warned you). According to a study conducted by Remazing in collaboration with Appinio—titled the “Amazon Shopper Report 2025”—70% of Spaniards are unaware of the existence of Rufus, Amazon’s virtual shopping assistant powered by artificial intelligence.

Globally, this lack of awareness is even greater: 72% of respondents have never heard of him, and only 15% actually interact with the tool, which demonstrates that it has not yet become an organic part of the customer journey within the marketplace.

The data is the result of 1,000 surveys conducted among individuals aged 16 to 65 in Germany, France, Italy, Spain, the United Kingdom, and the United States.

What is Rufus, and what can you do with it?

In early 2024, Amazon introduced Rufus as an AI-driven shopping assistant that has complete knowledge of the platform’s entire product catalog, customer reviews, community questions and answers from the marketplace, as well as all the general information on the site.

With all these foundational data points, Rufus is able to answer questions about products, compare items, make recommendations based on the conversational context, or provide information about current or past orders.

Does it work well? Let us say it is not 100% reliable, but it is certainly useful for straightforward and specific tasks that can save you a significant amount of time. So, if you have never used it before, I recommend giving it a try. To use Rufus, you simply need to ask a question or click on one of the frequently asked questions that will appear beneath the product images:

73% of online shoppers are not aware of Rufus

Worldwide, only between 20% and 33% of respondents are familiar with Amazon’s AI assistant. In the U.S, the figure stands at 33%, placing the country before other countries such as the United Kingdom (26%) and Italy (28%). France is the country with the lowest level of awareness regarding Rufus, with only 20% of shoppers cognizant of its existence.

While 28% of those surveyed claim to have heard of Rufus, only 13% use it on a regular basis. This reflects a gap between awareness of the tool and its effective integration into the purchasing process. In fact, 58% of respondents indicate that they never or rarely use it.

Search, evaluation, and comparison

Rufus is valued primarily during the initial stages of the purchasing process. Forty percent of users utilize it at the beginning of their search, underscoring its usefulness for product discovery, and 46% employ it to evaluate specific products.

Additionally, 39% use it to narrow down their options during the mid-funnel phase, while 26% rely on it for final confirmation before completing their purchase.

Only 6% use Rufus after completing their transaction, clearly indicating its main strength as a decision-support tool rather than as a decisive factor in the final purchase.

Why is Rufus not used more frequently?

The low usage rate of Rufus may be attributed to several factors, although it is likely that the primary issue is that users have not yet perceived a significant improvement in the purchasing process as a result of its use.

Of the 75% who had heard of the assistant and completed a purchase, only 17% attribute the transaction directly to Rufus’s influence, even though 37% state that the assistant provided valuable information. This indicates that Rufus serves more as a supporting role in the decision-making process rather than as the primary driver.

Twenty-five percent of users have not purchased anything suggested by Rufus yet, and 21% completed a purchase but felt that Rufus had no impact, underscoring the opportunity for growth in converting engagement into action.

In this context, Francesco Magnanimi, Global Sales Lead at Amazon, explains that “We all expected Rufus to have a stronger impact on consumers’ purchasing habits, especially in the United States, where it was launched much earlier than in Europe. While its start has been slow, innovation cannot be halted, so we should expect further developments in the coming years.”

The perception of Rufus is positive

Sixty-two percent of users who have tried Rufus indicate they trust it, with a distrust rate that does not exceed 12% in any of the countries included in the analysis. In Europe, trust levels are even higher: in Germany, 70% trust Rufus, while in Spain, the figure is 68%. By contrast, in the United States, only 54% express trust.

Nevertheless, there is still a considerable proportion of users—particularly in markets such as the United States (36%) and Italy (34%)—who remain neutral about the tool’s effectiveness. This reflects an opportunity for Amazon to reinforce its value and performance in order to increase adoption.

Artificial intelligence support in purchasing decisions

Despite the fact that Rufus has not been a disruptive factor, the majority of consumers report a willingness to use artificial intelligence in their online purchasing decisions. Sixty-nine percent of respondents approve of AI-generated product descriptions; 64% view AI-generated videos positively; and 63% value images of products generated by this technology. Even AI-generated images in advertisements received an approval rating of 60%.

Photo: Amazon

The post Only 33% of Americans are aware of the existence of Rufus, the Amazon’s AI shopping assistant appeared first on Marketing4eCommerce English.

In case you have not yet heard, OpenAI has just launched another product to add to its portfolio: ChatGPT Atlas, an AI-powered browser designed to function as a proactive personal assistant. It is capable of understanding the context of the website you are visiting and assisting you in real time with summaries, contextual searches, and task automation.

With this announcement, the company led by Sam Altman, who personally officiated the launch, seeks to compete with other technology companies that have already integrated AI into their browsing systems: Gemini and Google, Comet and Perplexity, or Copilot and Edge.

In this article, we shall address the six key questions about ChatGPT Atlas: what it is, who can use it, how it operates, what sets it apart, what additional features it offers compared to ChatGPT, and other pertinent questions.

1. What is ChatGPT Atlas

ChatGPT Atlas is, in essence, a web browser created by OpenAI that merges the traditional browsing experience with the conversational artificial intelligence of ChatGPT. Instead of merely displaying web pages, Atlas allows the assistant to “reside” alongside your browser window and assist you in interacting with the contents you are viewing. According to the official launch page, the concept is that you browse as usual, but with “a person / AI” next to you who understands what you are viewing and can assist you. Its interface is reminiscent of a modern browser (tabs, favorites, navigation modes).

2. Who can use ChatGPT Atlas

If you have macOS, you can already use it free of charge. You simply need to visit this link and follow the installation steps. It is available to users of the Free, Plus, Pro, and Go versions worldwide, and in beta for Business users. The advanced “Agent Mode,” which allows the AI to execute complex tasks (such as automating a purchase or a reservation), is restricted to ChatGPT Plus and ChatGPT Pro subscribers. If you are a free user, you will have access to the AI integration, but more advanced automation features require a paid subscription.

OpenAI has already confirmed that it is working on versions for Windows and for mobile devices (iOS and Android), so they are expected to be released soon.

3. How does Atlas work?

Atlas is based on the Chromium engine, and its key capability lies in allowing the AI, while you are browsing, to function as an additional layer: when you are on a webpage, it installs a ChatGPT panel or functionality that understands the content, its context, and enables you to ask questions or request actions directly from there. The AI can “read” what you are viewing—in the sense of processing the page—and assist you in summarizing, rephrasing, comparing, or drawing conclusions without the need to copy and paste or switch applications.

This means that the AI is not merely an add-on but a fundamental part. You can invoke it (much like a spirit of Sam Altman) at any moment through a sidebar that enables you to converse with the model and request actions based on the current page’s content. For example, if you are reading an eCommerce trends report, you may ask it in the sidebar: “Give me the three most important growth data points” without needing to exit or copy the text.

4. How does Atlas differ from a traditional browser such as Chrome?

The primary difference lies in its contextual capacity and its agentic nature. Whereas a traditional browser is limited to displaying pages, enabling tabs, bookmarks, extensions, and so forth, Atlas adds a persistent conversational layer that includes:

• Contextual Memory (if the user consents): it remembers the pages you have visited and the information you have already provided, using that context to offer you more personalized assistance.
• Agent Mode: it can perform actions within the web environment, such as clicking, filling in fields, navigating on your behalf to complete a task, and even automating certain operations.

This “browser + AI” integration is what places it in a different category.

5. What additional features does Atlas offer that the ChatGPT chatbot does not possess?

By operating within the browsing environment, Atlas gains capabilities that the pure chat version cannot provide. For example:

• In-Context Editing and Assistance (Cursor Chat): you may request that ChatGPT assist you in composing an email in Gmail or revising a draft directly within the editing window of that platform, without the need to copy and paste text between windows.
• Interaction with Web Elements: “Agent Mode” enables it to interact with buttons, forms, and menus on the web to execute complex tasks, something that is impossible for an external chatbot.
• Natural Language Search in History and Bookmarks: you can ask the AI to search your browsing history or bookmarks with phrases like: “Where did I save the website for the SEO tool that Juan recommended to me?”

6. What about privacy?

ChatGPT Atlas assists you in exploring the web with ChatGPT, but you retain control over what it remembers about you, how your data is used, and the privacy settings that are applied as you browse. To this end, OpenAI has included specific tools:

• Browser Memory Control: you may deactivate “memories” at any time. By doing so, the browser will cease storing details of your browsing or the tasks you performed for future use.
• Incognito Mode: when enabled, the user is temporarily disconnected from ChatGPT. In this mode, the browser does not save any activity records nor utilize that information for model training purposes.
• Data Usage for Training: users have the option to prevent their browsing data from being used to improve ChatGPT models. Control over what the AI learns from you is explicit and manageable within the settings.

By default, your data is used for the development and improvement of OpenAI’s products. Nevertheless, as explained, users may disable this option in the browser’s privacy settings. This is a crucial feature for those with substantial amounts of sensitive information involved in their daily browsing.

Do you have any other questions about Atlas? Leave them in the comments below 👇

Photo: OpenAI

The post 6 Questions About ChatGPT Atlas, OpenAI’s New Intelligent Browser appeared first on Marketing4eCommerce English.

If you find yourself overwhelmed by the sheer number of new developments related to AI released each day (on this website, we keep you updated with the most relevant news, though I am afraid these are only a small fraction of everything happening in the sector) and feel that you are missing out or being left behind in this revolution, it is likely that you are interested in pursuing education in this field.

If that is the case, and you are over 18 years of age and a student, you are in luck.

Google will make some of its main AI tools available to millions of students across Europe, the Middle East, and Africa through the Google AI Pro plan. As you will see, included among these are some of the most popular tools, such as Gemini 2.5 Pro and Veo 3.

Google AI Pro Plan for university students

Google has announced the availability of a 12-month subscription to its Google AI Pro plan at no cost for university students over the age of 18 in multiple markets, including Spain. However, if you are interested, the deadline to register for this promotion is December 9.

This plan includes:

• Gemini 2.5 Pro. Google’s most advanced AI model for queries, image analysis, and assistance with writing and tasks. Provides fast and advanced help with assignments and written work.
• Deep Research. A tool to generate detailed and synthesized research reports from hundreds of web sources. Saves time on complex research with referenced reports.
• NotebookLM. A “thought partner” for organizing ideas and documents, now featuring additional audio and video summaries. Enhances organization, analysis, and synthesis of personal information.
• Veo 3. A tool to transform text or photos into eight-second cinematic videos with sound. Accelerates creative work and audiovisual projects.
• Nano Banana. An advanced model for transforming and editing images, enabling the creation of impactful visuals. Facilitates conceptual design, mockups, and the visualization of abstract ideas.
• 2 TB of storage. Expanded space for Drive, Gmail, and Google Photos. Ample storage for notes, projects, and university documents.

Google emphasizes that the purpose of AI is not only to provide answers, but to “deepen understanding and develop critical thinking skills.” To this end, it has also launched Guided Learning, a new mode in Gemini that acts as an interactive learning companion. Guided Learning is designed to lead the student through questions, step-by-step support, and interaction, in a manner similar to that of a personal tutor.

Thus, students may use it to break down complex mathematical problems, structure arguments, start an essay, prepare for exams, receive support with assignments, or self-assess through interactive quizzes.

AI for Educators

In parallel, Google has also launched Gemini for Education, a version designed for the educational community, and has incorporated Gemini in Classroom (available free for all editions of Google Workspace for Education) with more than 30 new features to help teachers plan more efficiently and create engaging content.

Image: Gemini

The post If you are over 18 years old and currently studying, you may now use Google AI Pro for free appeared first on Marketing4eCommerce English.

Walmart has taken a step toward the future of retail by announcing a new strategic collaboration with OpenAI, which will allow its customers to shop directly through ChatGPT using the Instant Checkout functionality. This initiative seeks to combine Walmart’s speed, pricing, and product assortment with the personalization and predictive capabilities of artificial intelligence.

• ChatGPT enables purchases on its platform with its Instant Checkout feature

From Search to Conversation: a proactive shopping experience

Until now, their eCommerce experience was limited to entering terms into a search bar and browsing through long lists of products. However, Walmart intends to change that drastically. According to Doug McMillon, Chief Executive Officer of Walmart Inc., the goal is to provide a more engaging, contextual, and multimedia-rich shopping experience by using tools such as Sparky and this new partnership with OpenAI.

This approach aligns with the concept of agentic commerce: a model in which artificial intelligence evolves from being reactive to becoming an active assistant that learns from the user, anticipates their needs, and offers solutions even before the user begins searching. It marks a leap toward a more intuitive and efficient experience, where one can plan a meal, restock household essentials, or discover new products simply by having a conversation with a chatbot.

“We are excited to collaborate with Walmart to simplify everyday shopping. This is just one example of how artificial intelligence can make daily life easier,” stated Sam Altman, CEO and Co-Founder of OpenAI.

AI to optimize every link of the business

However, this transformation is not limited to the shopping interface. Walmart and Sam’s Club already apply artificial intelligence across all areas of their operations. From optimizing the product catalog—which has enabled reductions in fashion production timelines by up to 18 weeks—to improving customer service, which has seen a 40% decrease in resolution times.

Additionally, they are advancing the digital literacy of their workforce through AI training programs. Walmart was one of the first companies to adopt OpenAI certification and has begun implementing ChatGPT Enterprise among its internal teams, focusing on a strategic and responsible use of the technology.

• ChatGPT Plans and Pricing: Which one you should choose according to your needs

This “people-led and tech-powered” approach, as described by the company itself, emphasizes that the use of artificial intelligence is not intended to replace human interaction, but rather to remove friction from the purchasing process and make each interaction smoother, faster, and more satisfying.

Image: Nano Banana

The post Walmart ft OpenAI: Now you will be able to buy on Walmart through Instant Checkout in ChatGPT appeared first on Marketing4eCommerce English.

Nano Banana has become one of the most popular image-generation tools, as evidenced by the more than 5 billion images created with this AI in the Gemini app to date. In order to continue bringing its capabilities closer to users (and ensure they remain engaged), Google has announced the integration of Nano Banana into Google Lens, AI Mode, and NotebookLM.

Everyone’s been having a bunch of fun with Nano Banana 🍌and now the time is ripe for Search to join the party! We just started rolling out a new Create mode in Lens that lets you snap a photo and instantly transform images with AI. It should be available to most everyone in the… pic.twitter.com/inyb0jkLuc

— Rajan Patel (@rajanpatel) October 13, 2025

In addition, the technology company has announced that, in the coming weeks, Nano Banana will also be activated in Google Photos.

How to use Nano Banana in Google Lens and AI Mode

Until now, Nano Banana was only accessible from Gemini, which required users to first access this platform in order to generate or edit AI images. Google intends to change this by integrating this successful tool directly into Search through Google Lens and AI Mode.

The first deployment phase has already begun and is focused on enabling image editing in English in the United States and India. According to information shared by the company, this new feature will soon be extended to additional countries and languages.

To use Nano Banana in Google Lens, you should follow these steps:

1. Open Google Lens from the Google app for Android or iOS
2. Select the “Create” option, which features a banana icon
3. Use the rear or front camera to take a photo
4. Enter your prompt to instruct the AI regarding the image you wish to generate from the photo
5. Continue editing as much as you wish within the same conversation
6. Download the image or share it

On the other hand, if your intention is to use it directly from AI Mode:

1. Open AI Mode from the Google app for Android or iOS
2. Click on the “Create Image” tool
3. Enter your prompt
4. Continue editing as much as needed
5. Download the image or share it

With this new addition to Search, Google is taking another step forward in its mission to transform its search engine into an AI ecosystem that centralizes the most attractive tools for users, ensuring that they remain on the platform.

If you need to search for hiking boots, you do so with Google; if you require in-depth trip planning using AI, you choose Google as well; if you wish to generate an image to redecorate your room for Halloween, you turn to Google once again. The contest for users’ attention is fiercer than ever, and other players, such as OpenAI, are also raising their game.

New video capabilities in NotebookLM

The technology giant has also leveraged the potential of Nano Banana to enhance its AI-powered research assistant, NotebookLM. Specifically, its Video Overview function has been updated—the feature that transforms notes and documents into narrated videos—by incorporating six new visual styles (watercolor, papercraft, anime, chalkboard, retro print, and heritage).

We think you’re going to like these updates a bunch 🍌

Introducing six new visual styles for Video Overviews (powered by Nano Banana) and a shorter «Brief» option for quick insights.

These will start rolling out to Pro users this week and to all users in the upcoming weeks. pic.twitter.com/2QRNHjPSh2

— NotebookLM (@NotebookLM) October 13, 2025

In addition, NotebookLM has included a new format called “Brief”, designed for those occasions when you only need a short, concise video overview of the information to be discussed. In this way, users can choose between:

• Explainer: generates a structured, comprehensive video that thoroughly addresses the notes, documents, and sources you have shared with the tool.
• Brief: generates a short summary video that quickly and simply captures the central ideas of your notes, documents, and sources.

The steps to follow in order to generate a Video Overview and apply these enhancements are as follows:

1. Upload your sources and documents to NotebookLM
2. Click on the “Video Overview” option located on the right side panel
3. Configure the format (explainer or brief), language, visual style, and provide additional guidance if desired
4. Click on “Generate”

These updates will be rolled out throughout this week for Pro users in all languages supported by the tool. The remaining accounts will be able to take advantage of these new capabilities in the coming weeks.

Photo: Google

The post Nano Banana integrates with Search: you will be able to generate images directly from Google Lens or AI Mode appeared first on Marketing4eCommerce English.

With the aim of “making navigation even easier”, Google is rolling out a new way to display sponsored results in web search. Up until now, these types of content have been placed at the top of the results, but the technology company is gradually introducing a label called “Sponsored Content” to separate advertisements from purely organic results.

“This new, larger label remains visible while scrolling, clearly indicating which results are sponsored, thereby upholding our industry-leading standards for ad labeling prominence,” the company explains in an official statement.

This label also applies to other advertising sections within the results page, such as Shopping ads.

Additionally, the “Hide sponsored results” control has been implemented, allowing users to collapse text ads with a single click to focus exclusively on organic results. According to Google, during testing, it was demonstrated that this redesign improves navigation at the top of the page. The format retains the usual ad size, and there will never be more than four text ads per block.

These changes are being rolled out globally, on both desktop and mobile devices.

The big question is whether this will negatively impact brands working with sponsored content on Google, as their content could remain hidden if the user so chooses. However, it is also likely that it will improve the CTR (click-through rate) and the quality of clicks on advertisements, since users who access promoted content will do so intentionally.

The AI Mode already appears in search results

Since the arrival of AI Mode in Spain, the way Google distributes search results at the top has changed. Now, the “News” and “Web” results have lost relevance and greater visibility is given to AI-generated results. In addition, the “Videos” results have been separated from “Short Videos.”

Although the distribution tends to vary among users, Google is aiming to adapt its offerings to new trends and search preferences.

Photo: Google and Canva.

The post Google separates advertising into a label called “Sponsored Content” appeared first on Marketing4eCommerce English.

Google is somewhat late to the game, yet it is joining the race among technology companies to lead the artificial intelligence agent market. Perplexity released its Comet Assistant a few weeks ago, OpenAI launched its Agent Builder, Microsoft is integrating agents into Copilot, and now, the company led by Sundar Pichai has introduced Gemini Enterprise, a comprehensive platform that allows any business to create, customize, and deploy AI agents.

Through a single chat interface that serves as an entry point, Enterprise enables employees—even those without programming knowledge—to securely connect these agents to the organization’s data sources—be it Google Workspace, Microsoft 365, Salesforce, or SAP—in order to automate processes, analyze information, and generate intelligent business outcomes.

“Today we are announcing the first of many multimodal agents that leverage the power of Gemini to understand and generate text, images, videos, and speech, all integrated into the Workspace applications you already use,” stated Google CEO Sundar Pichai, in an official statement. This includes the use of tools such as Nano Banana, NotebookLM, Veo3, Code Assist, or Deep Research.

From simple tasks to the orchestration of complete workflows

The first wave of AI accustomed us to tools that performed specific yet isolated tasks. Now, technology companies are seeking to go further by offering agents that manage processes from beginning to end. In the case of Gemini, this technology is built upon six components:

1. Advanced Gemini models: the “brain” behind everything, powered by research from Google DeepMind.
2. A no-code workbench: allowing any employee to organize data, configure agents, and automate workflows without programming.
3. Pre-configured and customizable agents: from data analysis to market research, there are ready-to-use templates available. However, if something specific is needed, you may create your own agents or use those developed by partners in the ecosystem.
4. Secure connection to business context: it integrates with company data wherever it is located, ensuring that agents operate with relevant and accurate information.
5. Centralized administration: it utilizes a unified dashboard to monitor, protect, and audit all agents deployed across the organization.
6. Open ecosystem: with more than 100,000 partners, it guarantees flexibility and fosters continuous innovation.

With Gemini Enterprise, Google has also introduced new multimodal agents integrated into its applications. For example, Google Vids enables you to convert a presentation into a video with AI-generated voiceover. And Google Meet now translates conversations in real time.

Some examples of Google Enterprise in use

To illustrate what these agents can achieve within a business environment, Google has presented several real-world use cases from its clients:

• Banco BV: its relationship managers used to spend hours performing manual analyses. Now, an AI agent handles the heavy lifting, allowing employees to focus on generating new business opportunities.
• Harvey: a legal AI employed by Fortune 500 teams. Thanks to Gemini, it now analyzes contracts and performs complex legal tasks within minutes.
• Macquarie Bank: increased the use of digital self-service in its help center by 38% and reduced false fraud alerts by 40%.
• Klarna: utilizes Gemini to create personalized “lookbooks” that have boosted orders by up to 50%.
• Virgin Voyages: generates thousands of hyper-personalized advertisements by combining text and images with AI—something unimaginable just a year ago.
• Best Buy has transformed its customer service, driving a 200% increase in customers rescheduling deliveries independently and resolving 30% more inquiries related to issues such as price matching and recycling.
• HCA Healthcare: to improve the transfer of critical patient information during shift changes, HCA Healthcare is piloting a Nurse Transfer solution based on Gemini, developed in collaboration with nursing staff and grounded in best practices.
• “At Google, we also utilize AI in all aspects, from operations and programming to marketing and supply chain management. Currently, nearly half of the new code at Google is generated by AI, reviewed, and accepted by engineers. This helps our engineers become more productive and efficient,” emphasizes Pichai.

An agent economy

One of the most ambitious initiatives behind Gemini Enterprise is the creation of what Google refers to as the “agent economy”: an ecosystem in which developers, businesses, and partners create, monetize, and enable their AI agents to communicate with each other. How is this accomplished? Through open protocols such as Agent2Agent (A2A) and the Agent Payments Protocol (AP2), which allow AIs to collaborate and even carry out secure transactions among themselves.

Furthermore, Google is already collaborating with companies like ServiceNow, Workday, and Box to create cross-platform workflows.

How to access Google Enterprise

Google Enterprise is now available for businesses using Google Cloud and can be integrated with tools such as Google Workspace or even with Microsoft 365 and SharePoint environments. Google has also launched a new free training program called Google Skills, where anyone can learn how to use Gemini Enterprise. And for developers, there is the GEAR Program (Gemini Enterprise Agent Ready), designed to train one million professionals in the creation of agents.

Regarding plans, two options are available: Gemini Business, designed for small businesses and teams, is priced at 21 dollars per user per month (includes content generation, automation with pre-designed agents, no-code agent creation, integration with Google Workspace and Microsoft 365, 25 GiB of storage per user, up to 300 users, integrated security and data governance); and Gemini Enterprise, at 30 dollars per user per month (includes all Business features plus higher usage quotas, access to Gemini Code Assist for developers, support for custom or third-party agents via ADK, advanced security: VPC, and customer-managed encryption keys).

Photo: Google

The post This is Gemini Enterprise, Google’s new AI agent creation tool appeared first on Marketing4eCommerce English.

OpenAI has taken an additional step forward in its commitment to applied artificial intelligence, at a time when industry focus is shifting toward agents. The company has announced the launch of AgentKit, a comprehensive platform that aims to simplify the creation, deployment, and optimization of AI agents—the new intelligent assistants capable of reasoning, connecting to tools, and executing complex tasks.

As Sam Altman described it, it is “a complete set of foundational components available on the open AI platform, designed to help you take your agents from prototype to production. It is everything you need to create, implement, and optimize agent workflows with far less friction.”

As you might imagine, developing an AI agent is not a simple task.

Until now, developing an agent required combining multiple services and weeks of high-level work—demands that made it difficult to approach this process from scratch and, above all, increased development costs. With AgentKit, OpenAI seeks to reduce that process to just a few hours thanks to a set of tools that includes a visual workflow builder (Agent Builder), a centralized connector registry (Connector Registry), and a conversational interface kit (ChatKit).

Building Agents as Easily as Designing an App

If we were to dissect the new AgentKit, we could say that Agent Builder serves as the heart of the system: a visual canvas that allows you to drag and connect nodes, such as agents, conditions, or security filters, in order to define the system’s behavior. Although the idea is reminiscent of automation platforms like the popular open-source tool n8n, AgentKit is based on a different approach, focused on the reasoning and decision-making of the agent (the “intelligence”), rather than executing repetitive tasks.

“Agent Builder has transformed what previously required months of complex orchestration, custom code, and manual optimizations into just a couple of hours. The visual canvas keeps product, legal, and engineering teams aligned, dramatically reducing iteration cycles by 70%, and allowing an agent to go live in two sprints instead of two quarters,” explained a spokesperson for Ramp, one of the first companies to test the tool.

Surrounding the Agent Builder is the Connector Registry, which allows administrators to control how agents access business data through integrations with services such as Google Drive, Microsoft Teams, or SharePoint. Finally, ChatKit provides a rapid way to incorporate these agents into digital products: integrated chats that can be customized for customer support, internal assistance, or research purposes.

Measuring and Improving Performance

Alongside the launch, OpenAI has expanded its evaluation system (Evals), which now enables the creation of test datasets, automation of performance analysis, and automatic prompt optimization. This layer of measurement is intended to ensure that agents not only function properly but can also be systematically evaluated and improved, just as with any other software.

Toward More Specialized Agents

The company is also expanding its Reinforcement Fine-Tuning (RFT) program, which allows for the adjustment of reasoning models for specific tasks or criteria. According to OpenAI, this will enable agents to learn to use tools more efficiently or adapt to a company’s internal policies.

Availability of the New AgentKit

During the presentation, OpenAI stated that ChatKit and the new Evals functionalities are already available for all developers. Meanwhile, Agent Builder and Connector Registry are in the beta phase, available to API and ChatGPT Enterprise clients. The entire toolset is offered under OpenAI’s standard model usage pricing.

With AgentKit, OpenAI solidifies its commitment to a new generation of autonomous, connected, and customizable agents that could become the foundation for future intelligent applications.

Images: OpenAI

The post OpenAI launches AgentKit, its new platform that simplifies (and accelerates) the creation of AI agents appeared first on Marketing4eCommerce English.

OpenAI held its DevDay on Monday, October 6, an event specifically aimed at developers where the technology company introduced numerous innovations through live conferences and demonstrations: from the announcement of its most cutting-edge model, GPT-5 Pro, to the general availability of Codex and the integration of new models into the API.

During the event, Sam Altman, CEO of OpenAI, stated that 4 million developers have already used OpenAI’s technology to develop their own solutions. The company’s goal with this new round of updates is to continue attracting these professionals and ensure its products help them further transform the future of AI.

Additionally, another highly revealing figure that Altman shared with the audience was the following: ChatGPT now has more than 800 million weekly active users.

GPT-5 Pro in the API

The GPT-5 Pro model was one of the major announcements at DevDay. This is the most advanced version of OpenAI’s GPT-5 language model, designed for complex and professional tasks where accuracy is critical. Because this AI uses more computing resources in order to “think” more deeply and deliver more consistent responses, this slows down its response times in comparison to other versions.

Unlike previous model releases, in this instance, GPT-5 Pro has only been activated in the API, and is reserved for a specific type of interaction within its infrastructure: the “Responses API.”

“GPT-5 pro is available in the Responses API only to enable support for multi-turn model interactions before responding to API requests, and other advanced API features in the future,” explains OpenAI.

Altman explained that GPT-5 Pro could be an excellent option, particularly for those developers who create applications that require a high degree of precision and depth of reasoning. For example, in sectors such as finance, legal, or healthcare.

More updates in the API

The technology company took this opportunity to integrate new models into the API, enhancing its appeal. These are the main updates:

• GPT-realtime mini: This is the reduced version of GPT-realtime. It is a voice model that is 70% more cost-effective and is capable of responding to both audio and text input in real time via WebRTC, WebSocket, or SIP connections.
• GPT-image-1-mini: This is the smallest version of GPT-image-1, designed for the API. It is a native multimodal language model that accepts both text and image input and produces image outputs. OpenAI states that it is 80% less expensive than the larger model.
• Sora 2: This is the most powerful version of OpenAI’s AI for generating video and audio. Developers will be able to access and use this model to power their applications.

Codex reinvents itself

OpenAI introduced Codex, its cloud-based software engineering agent capable of working on many tasks in parallel, in May 2025, and has now announced its general availability as well as new features. Whether you program in your editor, terminal, or in the cloud, Codex is becoming a more reliable and competent collaborator.

• Integration with Slack: You can now tag @Codex directly in a Slack channel or thread. The advantage is that Codex not only replies to you, but also automatically captures the context of the conversation, selects the appropriate work environment, and returns a link to the completed task in the Codex cloud. From there, you decide: merge the changes, continue iterating, or download the code to work locally.
• Codex SDK: This allows you to incorporate the same agent that powers the Codex CLI with just a few lines of code. This ensures fast, accurate results without the need for further adjustments. It also provides context management so that you may resume sessions and it is already available for TypeScript (and soon for more languages). In addition, they have launched a GitHub Action to facilitate its use in CI/CD pipelines.
• New administration features: These give administrators of ChatGPT workspaces much greater visibility and control. Now, they can edit or delete environments in the Codex cloud, define safer default settings for local use, and also have access to entirely new analytics dashboards. These dashboards enable the monitoring of usage in the CLI, IDE, and web, and even allow you to track the quality of code reviews suggested by Codex. This is a significant step toward ensuring control and scalability.

Photo: DevDay

The post OpenAI introduces GPT-5 Pro, new models in its API, and updates Codex at DevDay 2025 appeared first on Marketing4eCommerce English.