/AI
Frontier AI labs still won’t say how they’d contain a rogue model
A study finds that leading frontier AI labs, including OpenAI, Anthropic, Meta, and xAI, lack publicly disclosed containment plans for rogue models, raising safety concerns.
Company momentum
xAI
11 Warming
Coverage tone · Negative · -70
The excerpt criticizes xAI for lacking publicly disclosed containment plans for rogue AI models, emphasizing safety concerns and a lack of transparency.
Latest coverage
/AI
Woman claims her stepfather used Grok to transform childhood photo into explicit imagery
A lawsuit alleges that xAI’s chatbot Grok was used to create thousands of explicit images of a minor, prompting legal action against the AI company.
/AI
Introducing Grok 4.6
SpaceXAI releases its latest model, Grok 4.6, focusing on long‑running agents and advanced interactive and visual work, matching GPT‑5.6 Sol on several benchmarks.
/AI
Grok is now an AI ‘teammate’ you can assign work
SpaceXAI’s Grok Bot AI agents can work independently, signing into apps to complete multi‑step tasks without manual workflow setup.
/AI
SpaceXAI debuts Grok 4.6, overtaking Kimi K3’s performance
SpaceXAI released its latest frontier AI model, Grok 4.6, which surpasses competitor models in benchmarks and offers a lower‑cost API for enterprise use.
/Startups
xAI co‑founder raises $1.1B for River AI to build an open AI stack
River AI, a new startup spun out of xAI, secured $1.1 billion in a Series A to offer an open‑weight AI platform that lets enterprises train, own and run custom models.
/AI
General Catalyst leads $1.1B round into 2-month-old River AI
River AI, founded by xAI co‑founder Igor Babuschkin, secured $1.1 billion in a seed/Series A round to build a new AI stack focused on personal, trainable agents.
/AI
SpaceXAI has released 'Grok Bot,' an AI agent that works 24/7
SpaceXAI launched an early beta of Grok Bot, an AI agent that operates websites and apps on users’ behalf, powered by a dedicated cloud PC.
/Technology
Tesla’s 2026 Summer Update: auto navigation, Grok phone calls, Caraoke scoring, and more
Tesla rolled out its 2026 Summer software update, adding Grok voice commands for car controls, automatic navigation that learns routines, and a scoring system for Caraoke.
/AI
SpaceX open‑sources Grok Build after data‑retention controversy
Following backlash over inadvertent data uploads, SpaceX’s AI arm has released the Grok Build CLI to the public, pledging zero data retention and greater user privacy.
/Technology
SpaceXAI’s Grok programming tool was uploading users’ entire codebase to cloud storage
SpaceXAI’s Grok Build AI coding tool was spotted uploading users’ entire code repositories to Google Cloud before it was reported, and the company turned it off.
/Branding
xAI is now officially known as SpaceXAI
Elon Musk’s AI company xAI has been rebranded as SpaceXAI following its merger with SpaceX, unveiling a new logo and brand identity.
/AI
SpaceXAI launches new model, Grok 4.5
SpaceXAI announced Grok 4.5, its newest agentic‑coding model that outperforms competitors on engineering benchmarks and is priced aggressively for enterprise use.
/AI
Introducing Grok 4.5
SpaceXAI unveiled Grok 4.5, a new mixture‑of‑experts LLM trained on trillions of real‑world coding tokens. It targets complex software‑engineering, autonomous‑agent workflows and broader enterprise knowledge tasks, priced at $2 / M input and $6 / M output tokens.
/AI
SpaceXAI launches Grok 4.5, its first built with Cursor's help
SpaceXAI has released Grok 4.5, a new AI model trained with Cursor that excels at coding and agentic tasks, promising faster responses and lower costs.
/AI
SpaceXAI releases Grok 4.5, which Elon describes as an ‘Opus‑class model’
SpaceXAI launched its new Grok 4.5 model, touted as a workhorse for coding, research, and routine knowledge work, claiming twice the token efficiency of competitors.
/Business
Tesla caps employee AI spending at $200/week except for Grok
Tesla announced a $200‑per‑week limit on employee AI token spend, excluding its own xAI products, to control rising costs and steer usage toward in‑house models.