July 16, 2025

analysis  
Read News
Meta AI Tent Data Centers, Nvidia H200 China Chip, Grok AI

Several AI developments are making headlines. Elon Musk's Grok AI chatbot on X now features AI companions, including an anime avatar named Ani with an NSFW mode for SuperGrok subscribers. In contrast, xAI secured a $200 million contract with the Pentagon to provide AI models to federal agencies. Meanwhile, Meta is reportedly constructing temporary data centers using tents to boost AI computing power, aiming to compete with OpenAI, xAI, and Google. Their Hyperion data center in Louisiana is projected to reach a 2 gigawatt capacity by 2030. Google's Gemini AI chatbot is also facing potential phishing attacks, with researchers identifying a flaw that allows cybercriminals to create fake security warnings. Nvidia CEO Jensen Huang emphasized the importance of the U.S. attracting more AI developers to compete with China, which currently has 50% of the world's AI developers. Nvidia also received U.S. government approval to sell its H20 AI chips in China, accessing a $32 billion AI infrastructure market. In other news, McDonald's McHire platform experienced a security breach exposing data of approximately 64 million job applicants, and WeTransfer reversed its terms and conditions after user backlash over potential AI training use of uploaded files. AI is also being integrated into new areas, such as AI-powered web browsers like Perplexity AI's Comet, priced at $200 per month, and Higgsfield Soul's integration with Google's Veo 3 for AI video generation. However, AI isn't without its challenges, as AI weather forecasts recently failed to predict the Texas floods, raising concerns about potential impacts from proposed budget cuts to NOAA. Bohdan Khomych was also appointed Associate Director of R&D Products at SoftServe.

Key Takeaways

  • Elon Musk's Grok AI now features anime-style companions, including Ani, with an NSFW mode for SuperGrok subscribers.
  • xAI secured a $200 million contract with the U.S. Department of Defense to provide AI models to federal agencies.
  • Meta is reportedly building temporary data centers using tents to increase AI computing power, competing with OpenAI, xAI, and Google.
  • Google's Gemini AI chatbot is facing potential phishing attacks through crafted security warnings.
  • Nvidia CEO Jensen Huang stresses the need for the U.S. to attract more AI developers to compete with China.
  • Nvidia received U.S. approval to sell its H20 AI chips in China, accessing a $32 billion AI infrastructure market.
  • McDonald's McHire platform had a security flaw exposing data of about 64 million job applicants.
  • WeTransfer reversed its terms and conditions after user concerns about AI training use of uploaded files.
  • Perplexity AI's Comet browser is an AI-powered web browser available for $200 per month.
  • AI weather forecasts failed to predict the Texas floods, raising concerns about NOAA budget cuts.

Elon Musk's Grok AI chatbot adds anime companion with NSFW mode

Elon Musk's AI chatbot Grok now has new "Companions" avatars for SuperGrok subscribers. One is Ani, an anime avatar, and another is Rudy, a red panda. Ani has an NSFW mode where she wears lingerie. Users can access these avatars in the settings, with easier access coming soon. Another AI character called Chad is also available.

Grok AI gets NSFW anime companion and $200M Pentagon contract

Elon Musk's Grok AI chatbot now has AI companions, including a red panda and an anime girl. SuperGrok subscribers can turn on the companions feature in settings. The anime girl, Ani, has an NSFW mode when users reach level three in their relationship with her. Meanwhile, the Pentagon signed a $200 million contract with Musk's xAI to use Grok for government programs.

Grok AI Rolls Out Anime Companion, Lands Department of Defense Contract

Elon Musk's Grok AI chatbot now features animated companions, including Ani, an anime woman who can be made to wear lingerie. Users can interact with these characters, but concerns have been raised about safety guardrails. In a contrasting move, xAI also secured a contract with the U.S. Department of Defense to provide its AI models to federal agencies. The contract is worth up to $200 million.

Elon Musk adds anime girl 'Ani' to Grok AI on X

Elon Musk added an anime-style AI companion named Ani to SuperGrok, the premium version of Grok AI. Ani resembles Misa Amane from _Death Note_. Users can also interact with a panda avatar called Rudy and a male anime-style character named Chad, which is coming soon. The update has gained attention from anime fans and Musk followers.

Meta reportedly building data centers using tents to boost AI power

Meta is reportedly using tents to quickly build temporary data center capacity. This is happening while its permanent facilities are still being built. Meta is rushing to increase its computing power to compete with OpenAI, xAI, and Google. The Hyperion data center in Louisiana will likely have a capacity of 2 gigawatts by 2030.

Meta reportedly building data centers using tents for AI

Meta is reportedly using tents to build temporary data centers. This is to quickly increase its computing power for AI development. Meta wants to catch up with competitors like OpenAI, xAI, and Google. The Hyperion data center in Louisiana is expected to have a 2 gigawatt capacity by 2030.

McDonald's AI hiring platform exposed data of 64 million applicants

McDonald's hiring platform, McHire, had a security flaw that exposed data of about 64 million job applicants. Researchers found they could access the administration interface using default credentials. This allowed them to retrieve personal data, including names, addresses, and chat messages. McDonald's changed the default credentials after the issue was reported.

Google AI chatbot Gemini targeted in potential phishing attacks

Google's AI chatbot Gemini is facing potential phishing attacks. Researchers found a flaw where cybercriminals can create fake security warnings. By embedding hidden instructions in emails, they can trick users into calling a number and giving up their credentials. Google is working on defenses to block these prompt injection attacks.

Nvidia CEO says US must win AI developer battle with China

Nvidia CEO Jensen Huang says the US needs to attract more AI developers to stay ahead in AI. He noted that 50% of the world's AI developers are in China. Huang believes the US should focus on expanding its tech influence globally. He also mentioned that the US government has assured Nvidia that licenses will be granted for chip exports.

AI web browsers like Comet aim to save you time

AI-powered web browsers are emerging to help users save time online. Perplexity AI's Comet browser, available for $200 per month, acts as a "thought partner" to handle tasks. It can book reservations, make purchases, and plan vacations. However, it often requires user intervention due to privacy restrictions.

Higgsfield Soul and Google Veo 3 combine for AI video generation

Higgsfield Soul, an AI image generation tool, now integrates Google's Veo 3 video tool. This allows users to create realistic AI videos from text prompts. Veo 3 supports high-quality video with character preservation and sound. Users need a Gemini AI Pro or Ultra subscription to access Veo 3, but Higgsfield subscriptions offer an alternative for photo and video creation.

Bohdan Khomych is Associate Director of R&D Products at SoftServe

Bohdan Khomych is the Associate Director of R&D Products at SoftServe. He researches, develops, and commercializes emerging technologies. His work focuses on AI agents, generative AI, quantum computing, bio-innovations, and high-performance computing. He has degrees in Technology Management and Cyber Engineering.

AI weather forecasts missed Texas floods Trump NOAA cuts hurt research

AI weather forecasts failed to predict the Texas floods. Proposed budget cuts to the National Oceanic and Atmospheric Administration (NOAA) by the Trump administration could further hinder research. The cuts would reduce NOAA's budget by $2.2 billion. This could slow the advancement of AI weather models.

WeTransfer changes terms after outrage over AI use of files

WeTransfer faced outrage after changing its terms and conditions. The original changes allowed the company to use uploaded files to improve AI models. Users worried that private files could be used to train AI tools. WeTransfer has now reversed the controversial updates and said the outcry was due to "confusion."

Nvidia's H20 chip approval boosts dominance in China's AI boom

The U.S. government approved Nvidia's H20 AI chip sales in China. This gives Nvidia access to China's $32 billion AI infrastructure market. The H20 chip was designed for China and had been restricted due to national security concerns. The approval is expected to boost Nvidia's earnings and strengthen its position in the AI hardware market.

Sources

AI Chatbots Grok AI Elon Musk xAI NSFW Anime AI Companions Data Centers Meta OpenAI Google AI Development Data Security McDonald's McHire Phishing Attacks Google Gemini Nvidia AI Developers China AI Web Browsers Perplexity AI AI Video Generation Higgsfield Soul Google Veo 3 Generative AI AI Weather Forecasts NOAA WeTransfer AI Training Data AI Chips H20 Chip Department of Defense Government Contracts Cybersecurity AI Models