Mundu osoko gobernuak Musken Grok estutzen ari dira, horrek sortu ditzakeen “deepfake” sexualengatik
Europan, hainbat gobernuk ikerketak abiatu dituzte horrelako edukiak legez kanpokoak diren ebaluatzeko.
Grok xAIk garatutako chatbot-a da Grok, Elon Musk aberatsaren jabetzakoa, eta horren inguruko polemika mundu osoan lehertu da azken egunotan, eduki sexual esplizituak sortzeko duen gaitasunagatik.
Besteak beste, Europako Batzordeak X-i, adimen artifizial hori integratuta dagoen sare sozialari, eskatu dio urte amaierara arte atxiki eta gorde ditzala Grokekin lotutako barne-dokumentu eta datu guztiak. Bitartean, Europako hainbat herrialdek mekanismo legalak aktibatu dituzte AA horrek sortutako "biluzien" sexu-irudien kezkaren aurrean.
Espainian, Datuak Babesteko Espainiako Agentziak (AEPD) agerian dauden eta ikusezinak diren arriskuez ohartarazi du, baita penalak ere, hirugarrenen irudiak argitaratzea dakartenak, benetakoak izan zein adimen artifizialeko sistemekin sortuak izan.
Erresuma Batuak ikerketa abiatu du eduki esplizitu horiek legez kanpokoak ote diren argitzeko. Bien bitartean, Frantziak Grokek sortutako irudi sexualak bidali dizkio fiskaltzari.
Alemanian, Wolfram Weimer Hedabide ministroak horrelako edukiei aurre egiteko neurri legalak hartzeko eskatu dio Europako Batzordeari, eta arazoa "sexu-jazarpenaren industrializazio" bihurtzeko arriskua dagoela ohartarazi du.
Italiako agintariek ohartarazi dutenez, baimenik gabe pertsona errealen "biluzien" irudiak sortzeko AA tresnak erabiltzea pribatutasunaren urraketa larriak izan daitezke, eta, kasu batzuetan, delitu penalak.
Polemika horren aurrean, xAi Grok-en garatzaileak dio legez kanpoko edukiaren aurka egiten ari dela, halakoak topatuz gero kendu egiten dituela, eta irudiak sortzeko eta editatzeko tresnen erabilera mugatu diela ordainpeko erabiltzaileei.
Muskek lehenago X-n adierazi zuen Grok eduki ilegalak sortzeko erabiltzen zuen edonork beste bide batzuetatik eduki ilegalak igoko balitu bezalako ondorioak jasango lituzkeela.
You might like
X will block Grok's creation of fair, sexualized images for all users
The social network has imposed further restrictions on artificial intelligence, following controversy over the creation of images of real people in intimate clothing, including minors. The measure also applies to paid subscribers.
ChatGPT and Copilot will stop operating on WhatsApp and only Meta AI will remain available
As a result of the new messaging platform policies, third party chatbots will be shut down and, as a result, OpenAI and Microsoft will be withdrawn from tomorrow.
Wikipedia turns 25 and the Basque edition becomes a reference for minority languages
The world's largest free and collaborative encyclopedia is celebrating its 25th anniversary with more than 60 million articles, while Wikipedia in Euskera stands out for its sustained growth, active community and support for the care and dissemination of the Basque language on the Internet.
They warn of the criminal dangers of publishing images of other people, whether true or created by Artificial Intelligence
The Spanish Data Protection Agency warns that the thoughtless use of these instruments may violate fundamental rights and give rise to legal liability, especially in cases of intimate, sexualized content or affecting minors.
"Your Year in ChatGPT", a new feature that summarizes interactions in 2025
The tool will be available to all users of both free and paid versions, provided that their storage memory and chat history are activated.
What's the point and how is the new + button from Google's search bar used?
At the left end of Google's search bar, the + icon has replaced the magnifying glass, allowing you to attach both an image and a file, and, aided by Artificial Intelligence, Google conducts an advanced search using these images and files as context.
Netflix to Warner Bros. He'll get Discovery for $82.7 billion
The two companies reached an agreement this Friday that includes film and television studios and HBO.
Cloudflare has fallen again worldwide in less than a month: numerous affected websites
This new mistake comes after another major global shutdown that took place a few weeks ago, affecting services like the X social network, the ChatGPT chatbot and the League of Legends video game.
Meta has been sentenced to pay $479 million to the digital press for unfair competition
Madrid's 15th Commercial Court has sentenced Meta to compensate 87 AMI publishers for gaining a competitive advantage through behavioral advertising on Facebook and Instagram.