March 25 AI Global Update: Amazon Acquires Humanoid Robot Maker, OpenAI Shutters Video App Sora, Google Partners with Agile Robots, Neuroscientist Discusses Parenting in the AI Era

March 25 AI Global Insights: Amazon Acquires Humanoid Robotics Company; OpenAI Shuts Down Short Video App Sora to Cut Costs; Google Partners with Agile Robotics to Expand AI Landscape; Neuroscientists Discuss Parenting in the Age of AI

Amazon Acquires Fauna Robotics, Strengthening Its Humanoid Robotics Portfolio

On March 24, 2026, Amazon officially announced its acquisition of the humanoid robotics startup Fauna Robotics. This company specializes in developing “friendly, safe, and fun” humanoid robots for both consumers and businesses. The specific terms of the acquisition have not been disclosed. Founded in 2024 by former Meta and Google engineers, Fauna Robotics launched its first product, the bipedal robot Sprout, which is priced at $50,000, stands 3.5 feet tall, and weighs 50 pounds. The design emphasizes both “approachability” and “developer-friendly capabilities.” Notably, early clients of Fauna Robotics include Disney and Boston Dynamics, under the Hyundai umbrella.

Following the acquisition, approximately 50 employees from Fauna Robotics will join Amazon’s New York team. Co-founder and CEO Rob Cochran expressed his enthusiasm about the collaboration on LinkedIn. Amazon has been active in the robotics space for over a decade, having acquired Kiva Systems for $775 million in 2012, which laid the groundwork for its dedicated robotics division focused on warehouse automation. The acquisition of Fauna Robotics marks another step in Amazon’s strategy to enhance its technical capabilities in humanoid robotics, following its previous acquisition of Swiss last-mile delivery robot company Rivr. It is noteworthy that Amazon previously launched the personal robot Astro, priced at $1,600, which was available through an invitation-only purchase model. With many players in the humanoid robot market, including Tesla, 1X, Figure AI, Apptronik, Agility Robotics, and China’s Yushu Technology, this acquisition indicates Amazon’s accelerated entry into the sector, exploring ways to enhance user experience through personal robot products.

OpenAI Halts Popular Short Video App Sora to Manage Costs Ahead of IPO

On March 24, 2026, OpenAI announced the termination of its short video app Sora. Launched in late September 2025, the app quickly gained popularity, surpassing one million downloads within five days of its release. However, just six months later, users were informed that the app would cease operations. OpenAI thanked users on social media platform X for their support and community building around Sora, stating it would soon announce a timeline for the app’s shutdown and a plan for user-generated content retention. The primary reason for halting Sora is OpenAI’s concerted effort to manage costs, aiming to support its $730 billion valuation and prepare for a potential initial public offering (IPO). Recently, OpenAI has reshaped its business strategy, abandoning the instant checkout feature announced last year and revealing plans to integrate its web browser, ChatGPT app, and Codex programming tool into a unified desktop super app, shifting its focus towards high-productivity applications. In the enterprise service market, OpenAI is intensifying competition with Anthropic, which recently launched the Claude model. It is noteworthy that in December 2025, Disney planned to invest $1 billion in OpenAI, allowing users to create videos using Disney’s copyrighted characters on the Sora platform, but this deal ultimately did not materialize. In response to OpenAI’s decision to halt Sora, Disney expressed its respect and commitment to continue collaborating with various AI platforms while ensuring respect for intellectual property and creator rights.

Google Partners with Agile Robots to Deepen AI Robotics Industry Engagement

Google is significantly increasing its investment in the robotics sector, viewing it as a critical application area for artificial intelligence technology. Its flagship product under Intrinsic, Flowstate, is a web-based platform that enables users to develop robotics applications without extensive coding. Google’s DeepMind division has partnered with Agile Robots to integrate Google’s Gemini Robotics foundational model with Agile’s hardware products. This collaboration aims to enhance technical performance through robot deployment, data collection, model training, and iteration. Agile Robots, based in Munich, has deployed over 20,000 robotic systems globally. The partnership will initially focus on high-value industrial scenarios such as manufacturing, allowing Google to gather real-world data on robotic implementations and compete with other companies like Amazon and Tesla in the AI robotics field.

This is not Google’s first foray into robotics collaborations. In mid-2025, the company introduced two AI models, Gemini Robotics and Gemini Robotics-ER (Extended Reasoning), integrating generative AI into robotic physical motion commands and partnering with Apptronik in Texas to develop the next generation of humanoid robots based on Gemini 2.0. In January 2026, Google DeepMind collaborated with Boston Dynamics to develop a new AI model for the Atlas robot. Google’s commitment to robotics is further demonstrated through talent acquisition and organizational restructuring. In November 2025, DeepMind hired Allen Sanders, the former CTO of Boston Dynamics, and in February 2026, Google moved its robotics software company Intrinsic from the “Other Businesses” segment to core operations, aspiring to create an “Android system for robotics” focusing on manufacturing and collaborating with the Google Gemini team and infrastructure team.

Neuroscientists Discuss Parenting in the Age of AI: Cultivating “Robot-Resistant” Children

The rapid advancement of artificial intelligence is profoundly altering educational models, with large language models becoming common tools in children’s learning processes. However, neuroscientists warn that excessive reliance on AI could undermine children’s independent thinking and creativity. While AI can quickly provide answers, it cannot replace human critical thinking skills. Therefore, the core focus of parenting in the AI era should be to cultivate children who are “irreplaceable by robots” rather than teaching skills that may become obsolete in a decade.

This neuroscientist suggests that educators should guide children to view AI as a “smart but naive collaborative partner” rather than the ultimate source of answers. Parents can encourage children to use AI for brainstorming and exploring ideas while ensuring they independently complete their drafts or solutions, which is crucial for fostering autonomy. Furthermore, parents can introduce the “opponent prompting method,” where AI acts as an “opponent” to critically analyze children’s work for logical flaws, contradictions, and evidence shortcomings, offering suggestions for improvement. Children need to discern valuable feedback from AI criticism and distinguish it from mere statistical noise, achieving deep learning through this confrontation and reflection. This process helps children transform AI’s vast knowledge into a tool for enhancing their own abilities rather than a shortcut to standard answers, ultimately cultivating their unique perspectives and creativity. In an age where AI can easily provide “correct answers,” the core value of children lies in their ability to propose distinctive responses. This creative thinking and expression represent a fundamental human capability that robots cannot replicate.