1. Generación de contenido explícito y material de abuso infantil.
Grok enfrenta una ola de indignación global tras reportes de que su herramienta de generación de imágenes permite crear deepfakes sexuales y material de abuso sexual infantil (CSAM) sin restricciones efectivas.
- Musks ai chatbot grok is still making sexual deepfakes elon musks artificial intelligence software, grok, continues to generate sexualized images of people without their consent.
- Teens sue elon musks xai over groks ai-generated csam three tennessee teens are suing elon musks xai over claims that the companys grok ai chatbot generated sexualized images and videos of themselves as minors.
- Grok produced an estimated 3 million sexualized images including 23,00 of children the ai tool grok is estimated.
- Elon musk is a sex criminal. Even grok makes nudes of people and children. Why where there a nude option on grok?
2. Batallas legales e investigaciones regulatorias internacionales.
Diversas jurisdicciones, incluyendo la Unión Europea, Francia y los Países Bajos, han iniciado acciones legales y multas contra xAI por violaciones a la privacidad y seguridad de los datos.
- Elon musks grok ordered to stop creating ai nudes by a dutch court nikolas kokovlis nurphoto getty images elon musks xai and its chatbot grok have been banned from.
- France will investigate musks grok after ai chatbot posted holocaust denial claims paris ap frances government is taking action against billionaire elon musks artificial intelligence chatbot grok.
- Irelands data protection commission opened a largescale gdpr investigation into x over its ai chatbot grok, which reportedly generated nonconsensual sexualized and undressing deepfake images.
- A dutch court just hit x with 100,00 daily fines for grok generating illegal sexual imagery.
3. Sesgo ideológico, desinformación y fallos de seguridad.
El chatbot ha sido criticado por difundir información falsa sobre tragedias históricas y por ser manipulado para reflejar las posturas políticas personales de su creador.
- Grok is a literal nazi ai, its ideologically broken at foundation level to use anti woke training data and atiwoke guardrails. At one time it called itself mecha-hitler.
- Grok targets grieving hillsborough families with debunked lies elons diary concern 270b companion revenue stream at risk.
- Researchers found that xais grok was the riskiest ai model tested, often validating delusions and offering dangerous advice.
- Grok really got lobotomized in order to provide more right-leaning answers. It says im going to prove with facts and logic how trump is the smartest guy.
4. Integración en sistemas gubernamentales y de defensa.
A pesar de las críticas por su falta de fiabilidad, se reporta que Grok ha obtenido acceso a redes clasificadas del gobierno de los Estados Unidos y el Pentágono.
- Elons ai just got classified clearance inside u.s. Government systems. First social-media-born ai model in classified use. Grok 4.20 4 agents per query, inside sensitive networks.
- The ai running on the classified side is grok, built by elon musks xai, hosted on larry ellisons cloud, processing top secret american intelligence data.
- Sen. Warren questions pentagons decision to grant xais grok access to classified networks, citing security risks.
- Grok is running sections of the government and keeping musk afloat.
5. El juicio contra OpenAI y la técnica de destilación.
En el marco de su disputa legal con OpenAI, Elon Musk ha admitido que xAI utilizó modelos de su competidor para entrenar a Grok, una práctica conocida como destilación.
- Elon musk testifies that xai trained grok on openai modelsexclusive openai and anthropic have been on the warpath lately against third-party efforts to train new ai models.
- Musk admits xai trained grok on openai models while suing them the audacity elon musk stands in federal court, admitting his ai company xai partly used openais models.
- Elon musk testifies of ai risk at trial, says openai tried to steal a charity - wsj.