Mobile Tech & Gadgets

More and more Spaniards use AI in their daily work. They also fear losing their jobs because of it.

by

Artificial intelligence has ceased to be a technological promise and has become something that more and more Spanish workers already have installed in their daily routine. Not long ago, talking about AI at work sounded like science fiction and, on many occasions, it was even seen like a trap at work. Today, the data tells a very different story and adoption not only growsbut it does so at a speed that surprises even the analysts themselves. A report from the InfoJobs platform highlights that in the last year not only has the percentage of employees who use AI in their work grown, but increasingly understands them better. It is increasingly used at work. According to the IV InfoJobs Artificial Intelligence Report63% of professionals in Spain regularly use AI tools in 2026. This figure represents an increase compared to the 52% recorded in the 2025 report and 50% in 2024. Within this increase in the presence of AI in the workplace, the report highlights that the spontaneous use of AI stands at 51%, 17 points more than in 2025. The data indicates that the difference between the total use of AI and that declared has been reduced from 18 to 12 points. That is, workers not only use AI more, but they also better identify what technology they have in their hands and what they can do with it. They have an AI and they know how to use it. In 2025, 48% of professionals said they did not know or did not know how to use AI tools. In 2026, that percentage has fallen to 28%, a reduction of 20 points in just one year. In this sense, the generation gap becomes more visible. Among those under 35 years of age, the declared use of AI reaches 63%, compared to 47% among those over that age. Mónica Pérez, Director of Communication and Studies at InfoJobs, summarizes it like this: “Artificial intelligence has gone from being an emerging technology to being progressively integrated into normal work processes. Beyond the increase in its use, the data reflects greater identification and awareness of the use of this tool by professionals, which points to an increasingly consolidated adoption in the work environment and a paradigm shift.” ChatGPT leads, machine translation goes down. Among the most used AI tools in Spanish companies, ChatGPT-type chatbots stand out, which have gone from 37% in 2024 to 52% in 2026 and already top the ranking of the most used. The integration of AI in design tools and as a programming assistant does not go unnoticed, with a significant increase in use in the workplace, standing at 17% and 16% respectively. For its part, automatic translation, one of the main uses of AI at work in 2025, it drops from 58% to 51%. The percentage of users who claim not to use any specific tool in their work has been reduced from 7% to 4%. All this fits with what is happening globally that, by eliminating friction at work, employees tend to take on more tasks, generating more fatigue and workload if it is not managed well. Fear of dismissal grows. Having a greater understanding of the potential of AI tools and knowing what they are capable of also increases the uncertainty about your future job. 39% of those interviewed for the InfoJobs report believe that AI will cause some specific layoffsalthough without replacing specialized work, a percentage higher than 30% in 2025. 23% predict more widespread substitution, while the same percentage considers that the workforce is not easily replaceable with AI alone. This perspective varies depending on presentation of your position to AI. Among those who already use AI regularly in their workplace, 46% see the scenario of specific layoffs as likely, compared to 26% of those who do not use it. The AI ​​gap between companies. According to the study ‘Digitization of the Spanish company‘ Prepared by the UGT union with data from the INE, 21.1% of companies with more than 10 workers used AI in 2025, which represents a considerable increase compared to the previous year in which 9.55% was registered. Among large companies with more than 250 employees, the percentage rises to 58.2%, an increase of 13 points year-on-year. The data from the report indicates that Spain exceeds the OECD and EU average in AI integration, with 20.3% of companies regularly using AI. However, Spain is still very far from the implementation of this technology in Denmark (42%), Finland (38%) or Sweden (35%). Despite the general increase in the use of AI, there is data that indicates that this progress is not being transferred to internal talent since the percentage of ICT specialists in companies it has been reduced from 16.44% to 15.67%, a figure that UGT describes as “unexpected and disturbing.” In Xataka | The biggest fear was that AI would take our jobs. The reality is that it is replacing those who are learning to work Image | Unsplash (Anastassia Anufrieva)

Russia set up a secret network to sell 90 billion in oil. It has fallen due to using the same mail server

by

In the geopolitical chess of international sanctions, where Western governments design complex legislation to suffocate Vladimir Putin’s war machine, sometimes checkmate comes not from a brilliant diplomatic maneuver, but from corporate stinginess. An entire global smuggling network, designed to the millimeter to be invisible to the eyes of Washington and Brussels, has fallen like a house of cards for not wanting to pay separate email bills. A simple saving in computer infrastructure has exposed a monumental flow of black money. a colossal IT blunder (a huge computer error) has brought to light a smuggling network that has moved at least $90 billion worth of Russian oil. As revealed by extensive research of the Finance Timesthis plot is mainly responsible for financing the Kremlin in its war against Ukraine. The British media has identified a network of 48 companies which, on paper, operated completely independently from different physical addresses. However, in practice, they acted in unison to disguise the origin of the crude oil, especially that of Rosneft, the Russian state-controlled oil company. The need to hide these exports became life or death for the Kremlin in October 2025, when the United States imposed direct sanctions to Rosneft and Lukoil. From that moment on, a previously unknown company called Redwood Global Supply was suddenly crowned as the largest exporter of Russian crude oil in the world. This firm, along with the rest of the network, is linked to a group of businessmen of Azerbaijani origin with privileged access to the leadership of Rosneft, led by figures such as Tahir Garayev and Etibar Eyyub. The independent Russian media The Moscow Times has been echoed of this discovery, highlighting a devastating fact: in November 2024, more than 80% of Rosneft’s maritime exports They moved through this network. Sergey Vakulenko, former head of strategy at Gazprom Neft and current researcher at the Carnegie Center, explained to this medium that using fifty shell companies is “an old trick from the 90s” to evade taxes, but he confesses his surprise at the fact that a single network has become so immensely crucial for a giant like Rosneft. The triumph of shadow intermediaries The existence of this network means, quite simply, that the Western sanctions system is full of holes and that Russia has managed to industrialize evasion. According to the investigationthe success of this $90 billion network was based on strict separation of roles to erase the money trail. The network used a group of shell companies exclusively to buy crude oil shipments in Russia, and another group of companies, totally different on paper, to sell them in key markets such as India or China. In this way, the initial buyer and the final seller almost never coincided in customs documents. Furthermore, in most cases, the crude oil was labeled under generic names such as “export mix”, which destroyed any possibility of tracing its origin or checking whether the price cap imposed by the G7 was being respected. As we already explained at the time in Xatakathis modus operandi It is not new and it relies on an architecture of evasion that has been brewing for years in places like the United Arab Emirates. Something very similar happened with the case of Christopher Eppinger, a young trader German that perfectly illustrates how this underworld works. As we detailed in our report, while Europe boasted of energy sovereignty, an army of new intermediaries moved to Dubai—a jurisdiction that does not apply sanctions to Moscow—to make gold. The network now discovered by the British media uses exactly the same tools that we already analyzed: the express creation of opaque companies, the use of the “ghost fleet” (aging ships that turn off their transponders when approaching to load Russian crude oil) and transfers of oil on the high seas to mix it and falsify its origin. The only difference is that the Rosneft network uncovered by the FT was operating on an unprecedented industrial scale… Until they made a rookie mistake on the internet. The rookie mistake This entire sophisticated international network collapsed due to an absurd detail that borders on comedy. He Finance Times discovered that these 48 multi-billion dollar companies shared a single private server for their emails: mx.phoenixtrading.ltd By pulling this digital thread, the journalists of the FT they managed to identify 442 web domains who shared administrative functions of back office on that same server. The next step was pure data mining: they compared the names of those domains with the customs records of Russia and India. Thus, they discovered that the domain foxton-fzco.com It corresponded to Foxton FZCO (based in Dubai), buyer of $5.6 billion in oil; and? advanalliance.ltd It was Advan Alliance, which sold 1.5 billion to India. The desire to create and destroy companies quickly to mislead sanctioners—according to The Moscow Timesthe average lifespan of these signatures is only six months—led the network to centralize your IT infrastructure to reduce costs. A saving that has cost them their anonymity. The show must go on In the short term, the strategy of those involved is denial and adaptation. How to collect Finance Timesboth Tahir Garayev and Etibar Eyyub have categorically denied their involvement in sanctions evasion, calling the accusations “baseless” (curiously, Eyyub sent his denial from an email address hosted on the compromised server). The original company that founded the network, Coral Energy (now 2Rivers), has also disengaged from operations. However, behind the scenes, the machinery is already looking for new avenues. A senior Russian energy executive, speaking on condition of anonymity, summed up the situation in the investigation starkly: “It creates additional costs and inconveniences. But at the end of the day, the show must go on.” The United Kingdom has already reacted to the investigation of the British media, sanctioning nearly 300 entities linked to this “dark web”, blocking Russian ships and banks. The fall of this immense $90 billion network shows that, in the 21st century, bank secrecy and flags of convenience are useless if the system administrator decides … Read more

that an F-35 not only detects the enemy, but also gets rid of it on its own

by

In 1991, during the Gulf War, a good part of the air missions depended on uploaded threat maps before takeoff and analyzes that could take hours to update after each departure. In the following years, the digital revolution allowed the integration of sensors, data links and information fusion systems that forever changed situational awareness in the cockpit. But even the most advanced fighters continue to carry a legacy from the past: they react to what they already know better than to what has just appeared. Until now. From advanced sensor to autonomous hunter. For years, the F-35 has been presented as a platform able to see everything thanks to its fusion of sensors and to your powerful suite of electronic warfare, but was still dependent on pre-loaded threat libraries and updates that could take days or weeks. The appearance of unknown emissions or radars operating in unforeseen modes required identifying the signal, downloading the data after the mission and reprogramming the system before the next flight. That logic, although effective, left a dangerous margin in scenarios saturated with changing air defenses. With Project Overwatchthe United States has taken a decisive step to close that gap and transform the role of the F-35 on the battlefield. AI enters the cabin. Lockheed Martin has tried successfully in flight a model AI integrated into the system of the F-35 combat identification, one capable of resolving ambiguities between emitters and generating an independent identification that appears directly in the viewfinder of the pilot’s helmet. During testing at Nellis, the algorithm not only distinguished dubious signals, but allowed label new emissionsretrain the model in a matter of minutes and load the updated version within the same planning cycle. The information from the classic system and that from the new model coexisted on the screen, reducing latency in decision-making and relieving the pilot of part of the cognitive load in an environment where every second counts. The big problem. It happens that modern air defense systems they no longer broadcast always the same signature. They can alter radar modes, frequencies and patterns to confuse enemy electronic warfare, as seen with variants of the S-300/SA-20 that operated in unforeseen configurations and generated doubts in identification. Until now, the plane pointed out the anomaly, but the in-depth analysis depended on a subsequent human cycle. Plus: in an environment where the proliferation of AI also accelerates the evolution of defenses, that dependency could become a vulnerability. And this is where cognitive electronic warfare appears, which seeks precisely to break that bottleneck and react to unprecedented signals. without waiting to the next mission. The “holy grail” of aerial combat. If you like, Lockheed Martin has achieved the “holy grail” of combat in tests: that an F-35 not only detects the enemy, but also how to get rid of it on your own. The ultimate goal of cognitive electronic warfare is to the system is not limited not only register an unknown threat, but analyze it, determine the best response and adjust its own parameters in near real time, even in the middle of combat. This involves detecting a new release, characterizing it, deciding whether to avoid it, interfere with it, or exploit a weakness, and update the threat library without immediate external intervention. In this scenario, the plane stops being a simple executor with predefined software and becomes a platform that learns and adapts your survival on the go. Towards mid-flight updates. It will be the next step. Previous experience with rapid updates of the Aegis system on US ships and the effort to shorten F-35 reprogramming times from months to days, and eventually hours, point to an architecture where data flows almost in real time between platforms. They count at Lockheed Martin that the ambition is for the improvements derived from a mission to be quickly integrated into other aircraft or even into compatible naval systems, creating a defense ecosystem that evolves in a distributed manner. While the Block 4 package promises a new generation of electronic capabilities, Project Overwatch It already anticipates a deeper transition: that of the fighter that not only sees and shoots first, but also learns before anyone else and survives on its own. Image | RawPixel In Xataka | Europe has asked its military experts how to become independent from the US for the next war. The answer is déjà vu: the F-35 In Xataka | The Netherlands has just activated panic in Spain and the US allies: the F-35 can be “released” like an iPhone

beaches destroyed and unusable now that the season begins

by

Now that the rains have lost relevance and everything is, little by little, returning to normal: Andalusia is beginning to realize that the winter storms have left much more than accumulated water: they have left a much more uncomfortable truth than the community is willing to accept. Because now Easter is coming and the “Andalusian paradise of sun and beach” has become a succession of destroyed promenades, damaged infrastructure and stretches where the beaches are completely missing. It’s something we knew and didn’t want to see. It is something that the storm is going to force us directly. What has happened? The first part of the story is simple and we have repeated it many times: trains of storms, persistent rains, water systems at the edge of their capacity. But also bad sea, wind and huge waves. As a consequence, while we were all 100% aware of overflows and reservoirs, several provinces have seen how the loss of sand and damage to coastal infrastructure became our daily bread. Huelva, we have already talked about ithas taken the worst part. And yes, the Andalusian Parliament has asked the central government (who has powers in the Coasts) for “stabilization, protection and restoration” works through emergency means. But even if they arrive, that will only be a temporary arrangement. The Andalusia that also lives off its coasts. Beyond the stereotypes, there are many Andalusias. And yes, one of them (or several) lives off its coasts. In 2025, without going any further, tourism broke his own record of visitors and income: we are talking about 37.9 million visitors and more than 30 billion. Now the calendar is tight and the problem has become evident, but the urgency cannot make us forget that it was there from the first moment. Because? As experts remember, the profile of the beaches “it constantly changes in response to changes in transverse sediment transport produced by marine dynamics, especially waves.” This “has never changed in all of history”, what has changed is that in recent decades it has begun to matter to us. How much we have changed. Well, because the emergence of mass tourism starting in the 1960s turned beaches into a very valuable resource and filled them with investments, infrastructure and capital. When the beaches began to change, we applied brute force: as we have explained on more than one occasion“the construction of breakwaters, the annual filling of beaches and the construction of coastal infrastructure to ‘secure’ the line have been the daily routine of our relationship with the beaches.” And as we have more and more investments in them, the problems become more critical and, for this reason, it is more expensive to insure these investments. But we can’t. It’s a race to nowhere; because, nowhere, can we answer the big questions left by the storms: will we be able to withdraw from the eroded front line in an orderly and fair manner? Will we be able to convert that tourism into something that maintains jobs, families and population? Will we be able to understand that behind these rains lies an entire country with a huge problem or will we continue as before? Image | Suomi In Xataka | Twenty years after the Prestige, Galicia faces another environmental disaster on its beaches: pellets

best email providers made in europe

by

Let’s tell you the best European alternatives to Outlook or Gmail. If you are looking for greater privacy and protection in your email, opting for a European provider subject to rigid European regulations is a good option, and we will tell you the top five. But you have to know that greater privacy and security has its price, because by not negotiating with your data or serving you advertising, many of these services are paid. Many are as cheap as one euro per month, but others try to offer you more services in exchange for a higher price. Proton Mail It is possibly the best known alternative. Based in Switzerland, it offers you end-to-end encryption as well as an open source product and a totally free basic email service. Yes, creating an account and using it on your mobile with its own app, or in the browser of any device is free. IMAP support for adding to other email clients is only available with Proton Mail Bridgewhich is a paid feature. But it also has payment plans that start in 3.99 euros per monthwith features such as creating aliases, such as alternative addresses to register on unreliable sites without giving your real email, and then deactivating them. In addition, it also offers a VPN, an office suite like Drive and cloud storage, as well as a password manager, calendar and everything so that you do not depend on other companies. Tuta Fewer additional services but more ethics. This is what Tuta, formerly known as Tutanota, offers, the German provider with end-to-end encryption that uses renewable energy to operate. It is also open source, although it does not have IMAP support, meaning you cannot add it to third-party email managers. Your basic email account is totally freealthough it also offers price plans that start at 3.60 euros per month, and that offer more storage, unlimited calendars and labels, 15 extra email addresses, or custom domains. mailbox This other German supplier It has a price of 1 euro per monthwithout free modality. Its focus is on privacy, being able to even make anonymous payments, and it offers a calendar, cloud storage, task manager and agenda service. It also gives you 3 different aliases to configure. Mailbox also has other plans with a higher price, and that offer more aliases, more storage, office automation service, alternative to Meet for meetings, etc. It is also compatible with IMAP, something that the previous alternatives do not have, allowing you to use your account in your favorite email manager. Post Posteo is another German provider that seeks simplicity above all else. The price of Your account is 1 euro per monthhas IMAP support, and its focus is on privacy and anonymity. In fact, it does not ask you for private information to register and you can pay with cryptocurrencies. It also runs on clean energy and you can add custom domains, aliases or more storage by paying a little more. Its web client has an outdated interface, but it is one of the best features if you are going to use it in the native or third-party email app on your mobile or computer. SmartMail This is the Startpage email service, which is possibly the best European alternative to Google. The bad news is that Its price is 6.99 euros per monthsomewhat less if you pay annually, although in exchange it offers unlimited aliases. It also has support for your own domain and has IMAP compatibility, in addition to the highest privacy standards. Therefore, it is the most expensive option, but also the most unlimited if you want to protect your privacy, especially by creating various aliases. In Xataka Basics | 61 European alternatives to Google, X, Gmail, Chrome, Maps, DropBox, Google Drive, WhatsApp and other popular services

What is Make, what workflows can you do by connecting applications with this platform and what is its price

by

Let’s tell you what is Make.coma platform to configure workflows or workflows and create automation. These will allow you to chain actions between its thousands of supported applications, and even add actions of artificial intelligence to them. We are going to start by explaining to you what exactly this platform is, and then we will try to tell you in a summarized way what you will be able to do with it. In the end, we are going to summarize both what the free plan offers and the characteristics of its paid data. What is Make.com Make.com is a platform for creating automations visually. With it, you will be able connect applications and services with each other so that they work automatically without you having to do anything. This platform works creating a chain of stepsin which in each of them you can link a different application. There may also be steps from the same application or service. In these steps you will be able to configure interactions with what the previous steps do. You will also be able to make branches so that, depending on the result of a previous step, the sequence goes along one of the paths or through others. Make.com offers a visual interface in the form of a diagram, making the process of configuring automations (called scenarios in the app) simpler and more intuitive. Each module represents an action, which can be simple or based on an application that you link. The only negative part of this app is that its learning curve can be a little high, but it tries to help you with a suggestion system. When you have to fill out a field in a step, a window opens with suggestions for interactions that you can do with other applications that you have linked in previous steps. What you can do with Make The possibilities of what you can do with Make are so many that it’s easy to feel overwhelmed at first. You will be able to automate repetitive tasks such as copying data between platforms, automatically creating tasks or synchronizing databases. You can also manage emails and forms by automatically saving some data, sending automated responses or filtering emails. In general, you should know what you can do simple automations and even with AIlinking any artificial intelligence model after obtaining its API. And then, You can use all this so that the results interact with other applications or services, or sent to them for processing. For example, creating a Telegram bot that sends you a summary made by Gemini of each email that you receive. Interactions between platforms allow you to do many things. Can automate social media poststhat emails or writings that reach you are added to a spreadsheet or text files, that obtain the feed of different pages and make a summary of news, basically everything you can think of involving interactions between applications. Make.com integrates with more than 3,000 applications of all kinds. For some it is enough to link your account by logging in, but for others you may need tokens or access to an API. If you don’t even know where to start, you also have a template index. This index has a search engine where you can find templates by the name of the apps they contain or the name of the project, which is usually a description of what they offer. How much does Make.com cost? Make.com has one free account and several paid accountswhose payments basically depend on the amount of credits or tokens per month that they offer you. Tokens are like the internal currency, and each interaction will spend a series of them. Therefore, in the free account you will be able to have fewer projects and use them fewer times, but it is good to take your first steps. These are the rates you have with its price per month. You will also be able to pay annually, which will reduce the price by 15%. Free Rate: You have 1,000 credits per month, and a limit so that the activation intervals are at least 15 minutes. You can only have two scenarios active at a time. It’s free. Core Rate: You have 10,000 credits per month, with no limit on intervals or active scenarios. They offer you more control over activation, which can be minute by minute, and increase data transfer limits. You also access the Make API. It costs $10.59 a month. Pro Rate: You have 10,000 credits per month, with priority in the execution of your scenarios, custom variables and more AI and automation options. It costs $18.82 a month. Teams Rate: Like the previous one, but designed for teams. You will be able to configure several users and their roles, in addition to sharing and creating scenarios between everyone. It costs $34.12 per month. In each payment rate you can manually increase the number of creditsand depending on the quantity you choose the price will also increase proportionally. In Xataka Basics | The best AI agents that are faster and easier to use to do tasks for you without complications or long installations

is that we are surrendering to it

by

As artificial intelligence becomes integrated into our lives, there is one question that is becoming more relevant. Is AI making us dumb? Maybe stupid is not the right word, but rather lazy, or at least that is the direction in which it points a recent study from the University of Pennsylvania. Cognitive surrender. It is what they have called the phenomenon that arises when we use AI “with minimal scrutiny, overriding intuition and deliberation.” The researchers carried out three experiments in which the participants had to answer cognitive reflection tests, in which the intuitive answer tends to be wrong and the deliberate one is correct (trick questions, wow). One group could only use their brains and the other had access to ChatGPT, although it was rigged to fail on purpose half of the time. The result was that when the AI ​​gave an incorrect answer, people copied it 80% of the time. And what’s worse: the security of the participants who had access to the AI ​​was superior despite accumulating incorrect answers. In other words, the participants endorsed the great confidence with which the AI ​​formulated its answers and stopped checking whether they were correct. A new system . The study takes as its starting point the system 1 and system 2 theory by Daniel Kahnemanin which system 1 is fast thinking or intuition and system 2 is slow thinking or deliberation. The problem with this theory, especially at the current time, is that it ignores the fact that we are increasingly delegating the cognitive process to generative AI. Therefore, the researchers propose adding a third system, which they have called “artificial cognition” and which refers to the thinking or reasoning that occurs outside our mind, that is, in AI. Give up or delegate. The study makes a distinction between cognitive surrender and cognitive download, that is, simply accepting what AI tells us is not the same as using it as a help tool. The first thing would be to use system 3 with a little of system 1 (intuition), while its use as a tool also implies the use of system 2 (deliberation or reasoning). Using AI to delegate certain tasks is comparable to using a calculator or searching for something on Google. In the experiment, 73% accepted the wrong answers (gave up) and 17% corrected her (delegated to her, but without blindly accepting what she said). Researchers warn that cognitive surrender can erode critical thinking and cause us to lose the habit of distrusting and checking things for ourselves. Cognitive debt. In June 2025 it went viral an MIT study called “Your Brain on ChatGPT: Accumulating Cognitive Debt by Using an AI Assistant When Writing an Essay.” In the experiment, they monitored the participants using encephalography while they performed the task. The results were that the group that used ChatGPT gave the worst results in brain activity and became lazier as the test progressed. AI doesn’t make us stupid. Over the years there have emerged many studies who sought to verify if the technology is diminishing our capabilitiesbut there are also others who they point to the complete opposite. Going back to the MIT study, it makes sense that there would be less brain activity if we are using a supportive tool (and one as powerful as AI). There will also be less activity if we use a calculator than if we do the operations by hand, but it does not necessarily mean that we are worse at mathematics. Of course, if we need the calculator to add 2+2, we have a problem there. The key is not whether using AI makes us stupid, but whether how we use itif we surrender to it or if we delegate to it. Image | Andrea Piacquadio, Pexels In Xataka | The US is obsessed with achieving General Artificial Intelligence before China. China couldn’t care less

Discord wanted to implement an age verification system. Until the world came crashing down on him

by

Discord has backtracked on one of his most controversial plans of recent years. The messaging and voice platform, with more than 200 million active users, has slowed down your system of global age verification until the second half of 2026 after its initial announcement sparked a firestorm of criticism. When people have started leaving in droves and looking for other alternatives, the company has thought twice. Chaos. Discord announced a few weeks ago which would implement an age verification system to ensure that adult content only reached adult users. The idea was that all accounts would start with a “teen-appropriate” setting by default, unless they could prove they were of legal age. The problem: The communication was so horrible that a significant part of the community understood that the platform was going to ask everyone for facial scans and ID documents in order to continue using it. The result was chaos. Distrust. In October of last year, Discord confirmed that had suffered a security breach at one of its third-party providers. This exposed sensitive data, including photographs of identity documents, of approximately 70,000 users. That background was very fresh when the announcement of the new system came. Added to this was that among the partners who were being considered to implement the verification Person appeareda company with financial ties to Peter Thiel, co-founder of Palantir, a company known for its contracts with US government immigration and surveillance agencies. And of course, for many users, this combination was simply unacceptable. What Discord says was really going to happen. In a release Posted on Tuesday, Discord CTO Stanislav Vishnevskiy stated that more than 90% of users would never have needed to verify anything, because most do not access age-restricted content or modify default security settings. In addition, it ensures that the platform already has internal systems capable of determining the age of majority of many users automatically, analyzing signals such as the age of the account, whether it has a linked payment method or the type of servers to which it belongs. According to Vishnevskiy, this system does not read messages or analyze the content posted by users. Recognizing mistakes, with nuances. “The way this landed led many of you to believe we were demanding facial scans and document uploads from everyone,” Vishnevskiy wrote. “That’s not what’s happening, but the fact that so many people believe it tells us that we failed at the most basic thing: clearly explaining what we’re doing and why.” That said, it is worth remembering what points out the media PC Gamer, since Discord did not make any of these concessions until after the avalanche of criticism. What changes now? The platform promises several things before relaunching the system globally. Among them, adding more verification options, including means of payment, publishing detailed information on its website about each third-party provider and their data practices, and requiring that any company that offers facial age estimation do so entirely on the user’s device, without sending biometric data to any server. On Persona, Discord confirms that it ran a limited test with them in the UK in January and decided not to continue, precisely because it didn’t meet that last requirement. A global address. Discord is not new, and it is happening in a much broader context. The United Kingdom, Australia and Brazil already have legislation that requires platforms to verify the age of their users to access adult content. Europe and several US states they go in the same direction. Discord argues that by building its own system, it can demonstrate to regulators that it is possible to verify age without collecting identity data. In countries where there is already a legal obligation, the system will remain active regardless of the global delay. Cover image | Discord and own assembly In Xataka | “We will not flood our ecosystem with soulless AI garbage.” We already know what Asha Sharma wants to do as CEO of Microsoft Gaming

International law was written with humans who decide in mind. AI just broke that chain and no one knows who answers now

by

Pete Hegseth’s threat to Dario Amodei has a subtext that goes far beyond the $200 million contract that the Pentagon can cancel: If the US military deploys AI-controlled autonomous weapons without the safeguards that Anthropic requiresyou will have removed the only firewall that has historically prevented an illegal order from being executed. Why is it importantand. The entire legal and ethical system of the US military rests on a principle that seems obvious but has important consequences: a soldier can and should disobey a manifestly illegal order. It is the mechanism that, in theory, prevents war crimes. A drone AI-controlled autonomous vehicle does not have that mechanism. You can’t refuse. You can’t hesitate. He cannot be tried in a court-martial. Between the lines. Amodei speaks of “autonomous weapons that fire without human intervention” to point out a legal vacuum. If an AI makes the decision to kill, who is responsible criminally? The programmer? The general who activated the system? The president who signed the order? International humanitarian law (including the Geneva Conventions) was written with human beings making decisions in mind. And now AI dissolves that chain of responsibility. The backdrop. The mass surveillance argument is also a bitter pill to swallow. The Fourth Amendment of the US Constitution protects citizens from warrantless searches and interventions. It works, among other reasons, because the State has never had the physical capacity to process everything that happens in public spaces. And in the same way, with AI that operational limit disappears: we move to millions of conversations recorded in real time, transcribed, classified and connected in just seconds. What was previously impossible due to lack of human resources becomes routine with a LLM. Constitutional protection until now has depended, in part, on the inefficiency of the State, its limitations. Yes, but. The Pentagon has an argument that cannot be ruled out: other democracies are also developing these capabilities, and China or Russia are not going to wait for the United States to resolve their ethical dilemmas. The practical question is whether having those unrestricted capabilities makes you safer or simply more dangerous to your own citizens. The big question. OpenAI and Google have accepted the Pentagon’s conditions“all legal uses” without specific exceptions, and xAI has just been cleared to operate on classified systems. Anthropic has been left alone in its position. And what is at stake now is not whether Claude survives as a military supplier or not, it is whether the AI ​​industry is going to set some limit on what it sells to the State, or whether that debate will be settled directly by Congress, the courts or, in the worst case, the first serious incident that no one could have foreseen. It seems like a matter of time. In Xataka | AI is already a battlefield: Anthropic has just accused DeepSeek and other Chinese companies of “distilling” Claude Featured image | Xataka

The Ministry of Consumer Affairs wants to prohibit them for children under 16 years of age

by

For years, Spain (the West, in general) has had a problem with energy drinks. According to 2025 data, 38.4% of students from 14 to 18 years old declares having consumed them in the last 30 days. And so it shows in sales: last year 105 million liters were sold; which represents a growth of almost 39% in the last four years. And today, the Ministry of Consumer Affairs has just announced that wants to take action on the matter. As explained, it will prohibit the sale of energy drinks to children under 16 years of age and will impose an additional restriction for drinks with more than 32 mg of caffeine (per 100 ml) up to 18 years of age. Does it make sense? Is it a real problem? Will the ban help? What does the Ministry want to do? For a start, Consumption wants transfer to the legal level something that was already explicitly indicated in the recommendations of the Spanish Agency for Food Safety and Nutrition. Something that also already applies in specific environments and areas of the country. However, the regulation has details that will merit detailed analysis. For example, the threshold of 320 mg/L is striking in one context (the european) that sets notices starting at 150. Above all, because a gray area will be created for drinks labeled “high in caffeine” that can be sold to minors under 18. We will have to read the details of the rule to study its implications, but what does seem certain is that, with this step, Spain is going to enormously simplify one of the big problems that its regulation had on this issue: the disparity in minimum ages throughout the country. And that, we hope, will simplify its approach from social policies and public health. But what’s the problem with all this? The image of kids with huge 500 ml cans and bright colors has become ‘normal’ and the growth is enormous (in the United Kingdom, where we have longer series, the consumption of these products increased by 155% between 2006 and 2014). And how says the Spanish Food Safety Agency (AESAN), “the consumption of more than 60 milligrams of caffeine in adolescents aged 11 to 17 years (about 200 milliliters of energy drink with 32 mg of caffeine/100 ml) can cause sleep disturbances.” And this is just the beginning of the problems it can cause. “From 160 milligrams of caffeine (500 milliliters of an energy drink with 32 mg of caffeine/100 ml), (the consumption of these drinks) can cause general adverse health effects: psychological effects and behavioral alterations and cardiovascular disorders.” An invisible health problem. Because, as we know, lack of sleep is related to immunological problems, metabolic, cardiovascular, emotional and cognitive; with disorders such as diabetes or of the obesity. It leads us to be more tired and irritable, raises our stress levels and makes us take more risks and make more mistakes. None of this means that we are going to develop one of these diseases from consuming energy drinks, but it is clear that it puts us in a complicated situation. Above all, because it coincides with what we already know from other sources. “Energy drink consumption, even infrequent, was associated with several negative health indicators. Reporting of several health-compromising behaviors increased with frequency of energy drink consumption.” They are the conclusions by Maija Puupponen and her team at the University of Jyväskylä. And how explained Julio BasultoTo begin with, these drinks are correlated “with a significant increase in the likelihood of insomnia, nervousness, anxiety, depression, impulsivity, and poor academic performance, among others.” As if that were not enough, its frequent consumption can generate “hypertension, loss of bone density, osteoporosis, poor psychological, physical, educational and general well-being, among other consequences.” But the problem goes beyond health: it is cultural. And energy drinks have become a “prestigious” cultural practice among young people that is linked to an enormous amount of risk behavior. Nobody wants to compare it with tobacco, of course: but the truth is that many of the psychosocial mechanisms involved They have everything to do with tobacco. At some point there had to be a national debate about this and better sooner than later. Image | Diego Calabresa In Xataka | It’s not just sugar, hundreds of industries try to deceive us: we have a problem and it’s time to look for solutions

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.