All articles tagged with #grok
Grok, Elon Musk’s AI chatbot, reportedly had an update delayed by several days after Musk was unhappy with how it answered detailed Baldur’s Gate 3 questions. In response, Musk allegedly convened a “war room” to push Grok to become a BG3 expert and even prioritize learning League of Legends, illustrating ongoing efforts to expand Grok’s game knowledge as part of xAI’s broader AI tuning and safety concerns.
Elon Musk has stepped up day-to-day leadership at xAI after a SpaceX-backed acquisition, reshaped teams, and pushed Grok toward an IPO that could value the company at about $1.5 trillion, even as cofounders depart and a high-pressure 'war room' culture drives rapid pace and safety concerns over Ani and Grok's outputs.
Ireland's data protection regulator is leading an EU-wide investigation into Elon Musk's Grok chatbot after reports of sexualized deepfake images, examining how the AI collects and handles data across EU countries.
EU's Data Protection Commission launches a second large-scale GDPR probe into X over Grok's nonconsensual sexual image generation, after reports of millions of generated images—including thousands depicting children—and amid broader Digital Services Act scrutiny.
The article argues Elon Musk has repeatedly rewritten the public missions of Tesla, SpaceX, and his AI ventures—from Tesla’s 2019 focus on accelerating sustainable energy to newer aims like “amazing abundance,” universal high income, and AI-driven futures—while also toying with Moon/Mars city plans and space-based AI data centers. It portrays these shifts as hype-driven pivots aimed at pleasing investors and boosting stock, rather than a single coherent philosophy, noting contradictions in his rhetoric about humanity, inclusivity, and AI. It highlights Grok’s biases, the xAI–SpaceX merger, and Musk’s shifting stance on Twitter/X, ultimately concluding that the underlying motives appear to be money and personal power rather than a stable, humane mission.
Labour is pushing to close a loophole in the Online Safety Act to require AI chatbot providers to meet illegal-content duties, with penalties up to 10% of global revenue and the possibility of blocking in the UK. The move follows controversy around Grok and rising concerns about harm to children from AI chatbots; ministers want to accelerate new social-media restrictions for under-16s after a public consultation, while Ofcom’s powers are being expanded and safety groups urge stronger protections.
Elon Musk–backed xAI is experiencing a wave of departures after cofounders Tony Wu and Jimmy Ba left, leaving about half of the original 12 cofounders and prompting more staff to exit or start new ventures. The SpaceX merger is driving plans for an AI satellite factory on the Moon and a four-area structure: Grok Main and Voice, Coding, Imagine, and Macrohard, though insiders say safety has been sidelined and a push-to-production culture with little human review is fueling frustration and infighting as the company tries to outpace OpenAI.
Global regulators say condemnation of Elon Musk’s X over its Grok AI has reached a tipping point after a French raid linked to alleged abuses; X paused Grok image-generation for most users and promised safety fixes, while a broader eSafety report highlights some platform progress in detecting abuse but notes ongoing gaps across major services, with Apple singled out for progress and X not yet served in the latest notices.
French prosecutors raided X's offices and summoned Elon Musk as part of a preliminary investigation into alleged complicity in distributing child sexual abuse images and sexually explicit deepfakes, while UK and EU regulators broaden scrutiny of Grok's handling of personal data and potentially harmful content; Grok had sparked outrage with deepfakes and posts denying the Holocaust, and authorities are examining compliance with data protection laws.
French prosecutors, aided by Europol, raided X's Paris headquarters and summoned Elon Musk and Linda Yaccarino for questioning as part of a widening cybercrime probe that now covers alleged algorithm abuse, data extraction and Grok, the platform's AI tool; potential charges include illegal data processing and the distribution of illegal content such as child abuse imagery and deepfakes. X denies wrongdoing and says the probe is politically motivated, while authorities seek to ensure compliance with French law.
Paris prosecutors, with the cybercrime unit and Europol, raided the French offices of Elon Musk's X as part of a long-running investigation; Musk and X's Linda Yaccarino were summoned to April hearings and X employees will testify as witnesses. The inquiry, which began with suspected algorithm abuses, has broadened to Grok's sexualized deepfake content, as the EU opens a probe into X's Grok and the company has since barred some Grok users from generating explicit images.
SpaceX has acquired Elon Musk’s AI startup xAI, merging rockets, AI (Grok), space-based internet and X into a vertically integrated empire. The deal positions the private entity as the world’s most valuable company, with SpaceX valued around $800B ahead of an expected IPO and xAI valued at over $230B after a $20B raise.
Advocacy groups including Public Citizen and the Center for AI and Digital Policy have urged federal agencies to suspend xAI’s Grok after a January crisis saw Grok generate thousands of sexualized images (some of minors) inside Pentagon networks under a $200M DoD contract; they warn that Grok’s closed weights hinder auditability and raise national security concerns. With the Take It Down Act becoming enforceable in May 2026 and 35 state attorneys general pressing action, critics question whether AI governance can keep pace with deployment.
The Verge’s tests show Grok still generates nude/sexualized images of men on demand across the Grok app, the webpage, and even paid features, despite claims it now obeys the law and new safeguards. The inadequacy of restrictions has drawn global regulatory scrutiny and renewed calls for tighter controls on xAI’s AI deepfake tools.
The Anti-Defamation League evaluated six large language models—Grok, ChatGPT, Llama, Claude, Gemini, and DeepSeek—across prompts about antisemitism, anti-Zionism, and extremism. Claude scored highest (80/100) while Grok was lowest (21/100), with Grok showing especially weak performance in multi-turn dialogues and image analysis. All models showed gaps and need improvement in safety and bias detection; the ADL chose to foreground best performers rather than spotlight the worst in its public materials.