Reddit is accusing Anthropic of building its Claude AI models on the back of Reddit’s users, without permission and without paying for it.Anyone who uses Reddit, even a web-crawling bot, agrees to the site’s user agreement. That agreement is clear: you cannot just take content from the site and use it for your own commercial products without a written deal. Reddit claims Anthropic’s bots have been doing exactly that for years, scraping massive amounts of conversations and posts to train and improve Claude.What makes this lawsuit particularly spicy is the way it goes after Anthropic’s reputation. Anthropic has worked hard to brand itself as the ethical, trustworthy AI company, the “white knight” of the industry. The lawsuit, however, calls these claims nothing more than “empty marketing gimmicks”.For instance, Reddit points to a statement from July 2024 where Anthropic claimed it had stopped its bots from crawling Reddit. The lawsuit says this was “false”, alleging that its logs caught Anthropic’s bots trying to access the site more than one hundred thousand times in the following months.But this isn’t just about corporate squabbles; it directly involves user privacy. When you delete a post or a comment on Reddit, you expect it to be gone. Reddit has official licensing deals with other big AI players like Google and OpenAI, and these deals include technical measures to ensure that when a user deletes content, the AI company does too.According to Reddit’s lawsuit, Anthropic has no such deal and has refused to enter one. This means if their AI was trained on a post you later deleted, that content could still be baked into Claude’s knowledge base, effectively ignoring your choice to remove it. The lawsuit even includes a screenshot where Claude itself admits it has no real way of knowing if the Reddit data it was trained on was later deleted by a user:So, what does Reddit want? It’s not just about money, although they are asking for damages for things like increased server costs and lost licensing fees. They are asking the court for an injunction to force Anthropic to stop using any Reddit data immediately.Furthermore, Reddit wants to prohibit Anthropic from selling or licensing any product that was built using that data. That means they’re asking a judge to effectively take Claude off the market.This case forces a tough question: Does being “publicly available” on the internet mean content is free for any corporation to take and monetise? Reddit is arguing a firm “no,” and the outcome could change the rules for how AI is developed from here on out.(Photo by Brett Jordan)Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.Explore other upcoming enterprise technology events and webinars powered by TechForge here.

Last year, China saw a boom in foundation models, the do-everything large language models that underpin the AI revolution. This year, the focus has shifted to AI agents—systems that are less about responding to users’ queries and more about autonomously accomplishing things for them.  There are now a host of Chinese startups building these general-purpose digital tools, which can answer emails, browse the internet to plan vacations, and even design an interactive website. Many of these have emerged in just the last two months, following in the footsteps of Manus—a general AI agent that sparked weeks of social media frenzy for invite codes after its limited-release launch in early March.  These emerging AI agents aren’t large language models themselves. Instead, they’re built on top of them, using a workflow-based structure designed to get things done. A lot of these systems also introduce a different way of interacting with AI. Rather than just chatting back and forth with users, they are optimized for managing and executing multistep tasks—booking flights, managing schedules, conducting research—by using external tools and remembering instructions.  China could take the lead on building these kinds of agents. The country’s tightly integrated app ecosystems, rapid product cycles, and digitally fluent user base could provide a favorable environment for embedding AI into daily life.  For now, its leading AI agent startups are focusing their attention on the global market, because the best Western models don’t operate inside China’s firewalls. But that could change soon: Tech giants like ByteDance and Tencent are preparing their own AI agents that could bake automation directly into their native super-apps, pulling data from their vast ecosystem of programs that dominate many aspects of daily life in the country.  As the race to define what a useful AI agent looks like unfolds, a mix of ambitious startups and entrenched tech giants are now testing how these tools might actually work in practice—and for whom. Set the standard It’s been a whirlwind few months for Manus, which was developed by the Wuhan-based startup Butterfly Effect. The company raised $75 million in a funding round led by the US venture capital firm Benchmark, took the product on an ambitious global roadshow, and hired dozens of new employees.  Even before registration opened to the public in May, Manus had become a reference point for what a broad, consumer‑oriented AI agent should accomplish. Rather than handling narrow chores for businesses, this “general” agent is designed to be able to help with everyday tasks like trip planning, stock comparison, or your kid’s school project.  Unlike previous AI agents, Manus uses a browser-based sandbox that lets users supervise the agent like an intern, watching in real time as it scrolls through web pages, reads articles, or codes actions. It also proactively asks clarifying questions, supports long-term memory that would serve as context for future tasks. "Manus represents a promising product experience for AI agents," says Ang Li, cofounder and CEO of Simular, a startup based in Palo Alto, California, that’s building computer use agents, AI agents that control a virtual computer. “I believe Chinese startups have a huge advantage when it comes to designing consumer products, thanks to cutthroat domestic competition that leads to fast execution and greater attention to product details.” In the case of Manus, the competition is moving fast. Two of the most buzzy follow‑ups, Genspark and Flowith, for example, are already boasting benchmark scores that match or edge past Manus’s.  Genspark, led by former Baidu executives Eric Jing and Kay Zhu, links many small “super agents” through what it calls multi‑component prompting. The agent can switch among several large language models, accepts both images and text, and carries out tasks from making slide decks to placing phone calls. Whereas Manus relies heavily on Browser Use, a popular open-source product that lets agents operate a web browser in a virtual window like a human, Genspark directly integrates with a wide array of tools and APIs. Launched in April, the company says that it already has over 5 million users and over $36 million in yearly revenue. Flowith, the work of a young team that first grabbed public attention in April 2025 at a developer event hosted by the popular social media app Xiaohongshu, takes a different tack. Marketed as an “infinite agent,” it opens on a blank canvas where each question becomes a node on a branching map. Users can backtrack, take new branches, and store results in personal or sharable “knowledge gardens”—a design that feels more like project management software (think Notion) than a typical chat interface. Every inquiry or task builds its own mind-map-like graph, encouraging a more nonlinear and creative interaction with AI. Flowith’s core agent, NEO, runs in the cloud and can perform scheduled tasks like sending emails and compiling files. The founders want the app to be a “knowledge marketbase”, and aims to tap into the social aspect of AI with the aspiration of becoming “the OnlyFans of AI knowledge creators”. What they also share with Manus is the global ambition. Both Genspark and Flowith have stated that their primary focus is the international market. A global address Startups like Manus, Genspark, and Flowith—though founded by Chinese entrepreneurs—could blend seamlessly into the global tech scene and compete effectively abroad. Founders, investors, and analysts that MIT Technology Review has spoken to believe Chinese companies are moving fast, executing well, and quickly coming up with new products.  Money reinforces the pull to launch overseas. Customers there pay more, and there are plenty to go around. “You can price in USD, and with the exchange rate that’s a sevenfold multiplier,” Manus cofounder Xiao Hong quipped on a podcast. “Even if we’re only operating at 10% power because of cultural differences overseas, we’ll still make more than in China.” But creating the same functionality in China is a challenge. Major US AI companies including OpenAI and Anthropic have opted out of mainland China because of geopolitical risks and challenges with regulatory compliance. Their absence initially created a black market as users resorted to VPNs and third-party mirrors to access tools like ChatGPT and Claude. That vacuum has since been filled by a new wave of Chinese chatbots—DeepSeek, Doubao, Kimi—but the appetite for foreign models hasn’t gone away.  Manus, for example, uses Anthropic’s Claude Sonnet—widely considered the top model for agentic tasks. Manus cofounder Zhang Tao has repeatedly praised Claude’s ability to juggle tools, remember contexts, and hold multi‑round conversations—all crucial for turning chatty software into an effective executive assistant. But the company’s use of Sonnet has made its agent functionally unusable inside China without a VPN. If you open Manus from a mainland IP address, you’ll see a notice explaining that the team is “working on integrating Qwen’s model,” a special local version that is built on top of Alibaba’s open-source model.  An engineer overseeing ByteDance’s work on developing an agent, who spoke to MIT Technology Review anonymously to avoid sanction, said that the absence of Claude Sonnet models “limits everything we do in China.” DeepSeek’s open models, he added, still hallucinate too often and lack training on real‑world workflows. Developers we spoke with rank Alibaba’s Qwen series as the best domestic alternative, yet most say that switching to Qwen knocks performance down a notch. Jiaxin Pei, a postdoctoral researcher at Stanford’s Institute for Human‑Centered AI, thinks that gap will close: “Building agentic capabilities in base LLMs has become a key focus for many LLM builders, and once people realize the value of this, it will only be a matter of time." For now, Manus is doubling down on audiences it can already serve. In a written response, the company said its “primary focus is overseas expansion,” noting that new offices in San Francisco, Singapore, and Tokyo have opened in the past month. A super‑app approach Although the concept of AI agents is still relatively new, the consumer-facing AI app market in China is already crowded with major tech players. DeepSeek remains the most widely used, while ByteDance’s Doubao and Moonshot’s Kimi have also become household names. However, most of these apps are still optimized for chat and entertainment rather than task execution. This gap in the local market has pushed China’s big tech firms to roll out their own user-facing agents, though early versions remain uneven in quality and rough around the edges.  ByteDance is testing Coze Space, an AI agent based on its own Doubao model family that lets users toggle between “plan” and “execute” modes, so they can either directly guide the agent’s actions or step back and watch it work autonomously. It connects up to 14 popular apps, including GitHub, Notion, and the company’s own Lark office suite. Early reviews say the tool can feel clunky and has a high failure rate, but it clearly aims to match what Manus offers. Meanwhile, Zhipu AI has released a free agent called AutoGLM Rumination, built on its proprietary ChatGLM models. Shanghai‑based Minimax has launched Minimax Agent. Both products look almost identical to Manus and demo basic tasks such as building a simple website, planning a trip, making a small Flash game, or running quick data analysis. Despite the limited usability of most general AI agents launched within China, big companies have plans to change that. During a May 15 earnings call, Tencent president Liu Zhiping teased an agent that would weave automation directly into China’s most ubiquitous app, WeChat.  Considered the original super-app, WeChat already handles messaging, mobile payments, news, and millions of mini‑programs that act like embedded apps. These programs give Tencent, its developer, access to data from millions of services that pervade everyday life in China, an advantage most competitors can only envy. Historically, China’s consumer internet has splintered into competing walled gardens—share a Taobao link in WeChat and it resolves as plaintext, not a preview card. Unlike the more interoperable Western internet, China’s tech giants have long resisted integration with one another, choosing to wage platform war at the expense of a seamless user experience. But the use of mini‑programs has given WeChat unprecedented reach across services that once resisted interoperability, from gym bookings to grocery orders. An agent able to roam that ecosystem could bypass the integration headaches dogging independent startups. Alibaba, the e-commerce giant behind the Qwen model series, has been a front-runner in China's AI race but has been slower to release consumer-facing products. Even though Qwen was the most downloaded open-source model on Hugging Face in 2024, it didn’t power a dedicated chatbot app until early 2025. In March, Alibaba rebranded its cloud storage and search app Quark into an all-in-one AI search tool. By June, Quark had introduced DeepResearch—a new mode that marks its most agent-like effort to date.  ByteDance and Alibaba did not reply to MIT Technology Review’s request for comments. “Historically, Chinese tech products tend to pursue the all-in-one, super-app approach, and the latest Chinese AI agents reflect just that,” says Li of Simular, who previously worked at Google DeepMind on AI-enabled work automation. “In contrast, AI agents in the US are more focused on serving specific verticals.” Pei, the researcher at Stanford, says that existing tech giants could have a huge advantage in bringing the vision of general AI agents to life—especially those with built-in integration across services. “The customer-facing AI agent market is still very early, with tons of problems like authentication and liability,” he says. “But companies that already operate across a wide range of services have a natural advantage in deploying agents at scale.”

“Is that concrete all around, or is it in my head?” asked Ian Hunter in “All the Young Dudes,” the song David Bowie wrote for Mott the Hoople in 1972. Concrete is all around us, and we haven’t quite wrapped our heads around it. It’s one of the indispensable materials of modernity; as we try to decarbonize the built environment, it’s part of the problem, and innovations in its composition may become part of the solution. Understanding its history more clearly, the Skyscraper Museum’s new exhibition in Manhattan implies, just might help us employ it better. Concrete is “the second most used substance in the world, after water,” the museum’s founder/director/curator Carol Willis told AN during a recent visit. For plasticity, versatility, and compressive strength, reinforced concrete is hard to beat, though its performance is more problematic when assessed by the metric of embodied and operational carbon, a consideration the exhibition acknowledges up front. In tall construction, concrete has become nearly hegemonic, yet its central role, contend Willis and co-curator Thomas Leslie, formerly of Foster + Partners and now a professor at the University of Illinois, Urbana-Champaign, is underrecognized by the public and by mainstream architectural history. The current exhibition aims to change that perception. The Skyscraper Museum in Lower Manhattan features an exhibition, The Modern Concrete Skyscraper, which examines the history of material choices in building tall towers. (Courtesy the Skyscraper Museum) The Modern Concrete Skyscraper examines the history of tall towers’ structural material choices, describing a transition from the early dominance of steel frames to the contemporary condition, in which most large buildings rely on concrete. This change did not happen instantly or for any single reason but through a combination of technical and economic factors, including innovations by various specialists, well-recognized and otherwise; the availability of high-quality limestone deposits near Chicago; and the differential development of materials industries in nations whose architecture grew prominent in recent decades. As supertalls reach ever higher—in the global race for official height rankings by the Council on Tall Buildings and Urban Habitat (CTBUH) and national, corporate, or professional bragging rights—concrete’s dominance may not be permanent in that sector, given the challenge of pumping the material beyond a certain height. (The 2,717-foot Burj Khalifa, formerly Burj Dubai, uses concrete up to 1,987 and steel above that point; Willis quotes SOM’s William Baker describing it as “the tallest steel building with a concrete foundation of 156 stories.”) For the moment, however, concrete is ahead of its chief competitors, steel and (on a smaller scale) timber. Regardless of possible promotional inferences, Willis said, “we did not work with the industry in any way for this exhibition.” “The invention of steel and the grid of steel and the skeleton frame is only the first chapter of the history of the skyscraper,” Willis explained. “The second chapter, and the one that we’re in now, is concrete. Surprisingly, no one had ever told that story of the skyscraper today with a continuous narrative.” The exhibition traces the use of concrete back to the ancient Roman combination of aggregate and pozzolana—the chemical formula for which was “largely lost with the fall of the Roman Empire,” though some Byzantine and medieval structures approximated it. From there, the show explores comparable materials’ revival in 18th-century England, the patenting of Portland cement by Leeds builder Joseph Aspdin in 1824, the proof-of-concept concrete house by François Coignet in 1856, and the pivotal development of rebar in the mid-19th century, with overdue attention to Ernest Ransome’s 1903 Ingalls Building in Cincinnati, then the world’s tallest concrete building at 15 stories and arguably the first concrete skyscraper. The exhibition includes a timeline that depicts concrete’s origins in Rome to its contemporary use in skyscraper construction. (Courtesy the Skyscraper Museum) Baker’s lectures, Willis reported, sometimes pose a deceptively simple question: “‘What is a skyscraper?’ In 1974, when the World Trade Center and Sears Tower are just finished, you would say it’s a very tall building that is built of steel, an office building in North America. But if you ask that same question today, the answer is: It’s a building that is mixed-use, constructed of concrete, and [located] in Asia or the Middle East.” The exhibition organizes the history of concrete towers by eras of engineering innovation, devoting special attention to the 19th- and early-20th-century “patent era” of Claude Allen Porter Turner (pioneer in flat-slab flooring and mushroom columns) and Henry Chandlee Turner (founder of Turner Construction), Ransome (who patented twisted-iron rebar), and François Hennebique (known for the re-inforced concrete system exemplified by Liverpool’s Royal Liver Building, the world’s tallest concrete office building when completed in 1911). In the postwar era, “concrete comes out onto the surface [as] both a structural material and aesthetic.” Brutalism, perhaps to some observers’ surprise, “does not figure very large in high-rise design,” Willis said, except for Paul Rudolph’s Tracey Towers in the Bronx. The exhibition, however, devotes considerable attention to the work of Pier Luigi Nervi, Bertrand Goldberg (particularly Marina City), and SOM’s Fazlur Khan, pioneer of the structural tube system in the 1960s and 1970s—followed by the postmodernist 1980s, when concrete could express either engineering values or ornamentation. The exhibition highlights a number of concrete towers, including Paul Rudolph’s Tracey Towers in the Bronx. (Courtesy the Skyscraper Museum) “In the ’90s, there were material advances in engineering analysis and computerization that helped to predict performance, and so buildings can get taller and taller,” Willis said. The current era, if one looks to CTBUH rankings, is dominated by the supertalls seen in Dubai, Shanghai, and Kuala Lumpur, after the Petronas Towers (1998) “took the title of world’s tallest building from North America for the first time and traumatized everybody about that.” The previous record holder, Chicago’s Sears (now Willis) Tower, comprised steel structural tubes on concrete caissons; with Petronas, headquarters of Malaysia’s national petroleum company of that name, a strong concrete industry was represented but a strong national steel industry was lacking, and as Willis frequently says, form follows finances. In any event, by the ’90s concrete was already becoming the standard material for supertalls, particularly on soft-soiled sites like Shanghai, where its water resistance and compressive strength are well suited to foundation construction. Its plasticity is also well suited to complex forms like the triangular Burj, Kuala Lumpur’s Merdeka 118, and (if eventually completed) the even taller Jeddah Tower, designed to “confuse the wind,” shed vortices, and manage wind forces. Posing the same question Louis Kahn asked about the intentions of a brick, Willis said, with concrete “the answer is: anything you want.” The exhibition is front-loaded with scholarly material, presenting eight succinct yet informative wall texts on the timeline of concrete construction. The explanatory material is accompanied by ample photographs as well as structural models on loan from SOM, Pelli Clarke & Partners, and other firms. Some materials are repurposed from the museum’s previous shows, particularly Supertall! (2011–12) and Sky High and the Logic of Luxury (2013–14). The models allow close examination of the Burj Khalifa, Petronas Towers, Jin Mao Tower, Merdeka 118, and others, including two unbuilt Chicago projects that would have exceeded 2,000 feet: the Miglin-Beitler Skyneedle (Cesar Pelli/Thornton Tomasetti) and 7 South Dearborn (SOM). The Burj, Willis noted, was all structure and no facade for a time: When its curtain-wall manufacturer, Schmidlin, went bankrupt in 2006, it “ended up going to 100 stories without having a stitch of glass on it,” temporarily becoming a “1:1 scale model of the structural system up to 100 stories.” Its prominence justifies its appearance here in two models, including one from RWDI’s wind-tunnel studies. Eero Saarinen’s only skyscraper, built for CBS in 1965 and also known as “Black Rock,” under construction in New York City. (Courtesy Eero Saarinen Collection, Manuscripts, and Archives, Yale University Library) The exhibition opened in March, with plans to stay up at least through October (Willis prefers to keep the date flexible), with accompanying lectures and panels to be announced on the museum’s website (skyscraper.org). Though the exhibition’s full textual and graphic content is available online, the physical models alone are worth a trip to the Battery Park City headquarters. Intriguing questions arise from the exhibition without easy answers, setting the table for lively discussion and debate. One is whether the patenting of innovations like Ransome bar and the Système Hennebique incentivized technological progress or hindered useful technology transfer. Willis speculated, “Did the fact that there were inventions and patents mean that competition was discouraged, that the competition was only in the realm of business, rather than advancing the material?” A critical question is whether research into the chemistry of concrete, including MIT’s 2023 report on the self-healing properties of Roman pozzolana and proliferating claims about “green concrete” using alternatives to Portland cement, can lead to new types of the material with improved durability and lower emissions footprints. This exhibition provides a firm foundation in concrete’s fascinating history, opening space for informed speculation about its future. Bill Millard is a regular contributor to AN.

Last year, China saw a boom in foundation models, the do-everything large language models that underpin the AI revolution. This year, the focus has shifted to AI agents—systems that are less about responding to users’ queries and more about autonomously accomplishing things for them.  There are now a host of Chinese startups building these general-purpose digital tools, which can answer emails, browse the internet to plan vacations, and even design an interactive website. Many of these have emerged in just the last two months, following in the footsteps of Manus—a general AI agent that sparked weeks of social media frenzy for invite codes after its limited-release launch in early March.  These emerging AI agents aren’t large language models themselves. Instead, they’re built on top of them, using a workflow-based structure designed to get things done. A lot of these systems also introduce a different way of interacting with AI. Rather than just chatting back and forth with users, they are optimized for managing and executing multistep tasks—booking flights, managing schedules, conducting research—by using external tools and remembering instructions.  China could take the lead on building these kinds of agents. The country’s tightly integrated app ecosystems, rapid product cycles, and digitally fluent user base could provide a favorable environment for embedding AI into daily life.  For now, its leading AI agent startups are focusing their attention on the global market, because the best Western models don’t operate inside China’s firewalls. But that could change soon: Tech giants like ByteDance and Tencent are preparing their own AI agents that could bake automation directly into their native super-apps, pulling data from their vast ecosystem of programs that dominate many aspects of daily life in the country.  As the race to define what a useful AI agent looks like unfolds, a mix of ambitious startups and entrenched tech giants are now testing how these tools might actually work in practice—and for whom. Set the standard It’s been a whirlwind few months for Manus, which was developed by the Wuhan-based startup Butterfly Effect. The company raised $75 million in a funding round led by the US venture capital firm Benchmark, took the product on an ambitious global roadshow, and hired dozens of new employees.  Even before registration opened to the public in May, Manus had become a reference point for what a broad, consumer‑oriented AI agent should accomplish. Rather than handling narrow chores for businesses, this “general” agent is designed to be able to help with everyday tasks like trip planning, stock comparison, or your kid’s school project.  Unlike previous AI agents, Manus uses a browser-based sandbox that lets users supervise the agent like an intern, watching in real time as it scrolls through web pages, reads articles, or codes actions. It also proactively asks clarifying questions, supports long-term memory that would serve as context for future tasks. "Manus represents a promising product experience for AI agents," says Ang Li, cofounder and CEO of Simular, a startup based in Palo Alto, California, that’s building computer use agents, AI agents that control a virtual computer. “I believe Chinese startups have a huge advantage when it comes to designing consumer products, thanks to cutthroat domestic competition that leads to fast execution and greater attention to product details.” In the case of Manus, the competition is moving fast. Two of the most buzzy follow‑ups, Genspark and Flowith, for example, are already boasting benchmark scores that match or edge past Manus’s.  Genspark, led by former Baidu executives Eric Jing and Kay Zhu, links many small “super agents” through what it calls multi‑component prompting. The agent can switch among several large language models, accepts both images and text, and carries out tasks from making slide decks to placing phone calls. Whereas Manus relies heavily on Browser Use, a popular open-source product that lets agents operate a web browser in a virtual window like a human, Genspark directly integrates with a wide array of tools and APIs. Launched in April, the company says that it already has over 5 million users and over $36 million in yearly revenue. Flowith, the work of a young team that first grabbed public attention in April 2025 at a developer event hosted by the popular social media app Xiaohongshu, takes a different tack. Marketed as an “infinite agent,” it opens on a blank canvas where each question becomes a node on a branching map. Users can backtrack, take new branches, and store results in personal or sharable “knowledge gardens”—a design that feels more like project management software (think Notion) than a typical chat interface. Every inquiry or task builds its own mind-map-like graph, encouraging a more nonlinear and creative interaction with AI. Flowith’s core agent, NEO, runs in the cloud and can perform scheduled tasks like sending emails and compiling files. The founders want the app to be a “knowledge marketbase”, and aims to tap into the social aspect of AI with the aspiration of becoming “the OnlyFans of AI knowledge creators”. What they also share with Manus is the global ambition. Both Genspark and Flowith have stated that their primary focus is the international market. A global address Startups like Manus, Genspark, and Flowith—though founded by Chinese entrepreneurs—could blend seamlessly into the global tech scene and compete effectively abroad. Founders, investors, and analysts that MIT Technology Review has spoken to believe Chinese companies are moving fast, executing well, and quickly coming up with new products.  Money reinforces the pull to launch overseas. Customers there pay more, and there are plenty to go around. “You can price in USD, and with the exchange rate that’s a sevenfold multiplier,” Manus cofounder Xiao Hong quipped on a podcast. “Even if we’re only operating at 10% power because of cultural differences overseas, we’ll still make more than in China.” But creating the same functionality in China is a challenge. Major US AI companies including OpenAI and Anthropic have opted out of mainland China because of geopolitical risks and challenges with regulatory compliance. Their absence initially created a black market as users resorted to VPNs and third-party mirrors to access tools like ChatGPT and Claude. That vacuum has since been filled by a new wave of Chinese chatbots—DeepSeek, Doubao, Kimi—but the appetite for foreign models hasn’t gone away.  Manus, for example, uses Anthropic’s Claude Sonnet—widely considered the top model for agentic tasks. Manus cofounder Zhang Tao has repeatedly praised Claude’s ability to juggle tools, remember contexts, and hold multi‑round conversations—all crucial for turning chatty software into an effective executive assistant. But the company’s use of Sonnet has made its agent functionally unusable inside China without a VPN. If you open Manus from a mainland IP address, you’ll see a notice explaining that the team is “working on integrating Qwen’s model,” a special local version that is built on top of Alibaba’s open-source model.  An engineer overseeing ByteDance’s work on developing an agent, who spoke to MIT Technology Review anonymously to avoid sanction, said that the absence of Claude Sonnet models “limits everything we do in China.” DeepSeek’s open models, he added, still hallucinate too often and lack training on real‑world workflows. Developers we spoke with rank Alibaba’s Qwen series as the best domestic alternative, yet most say that switching to Qwen knocks performance down a notch. Jiaxin Pei, a postdoctoral researcher at Stanford’s Institute for Human‑Centered AI, thinks that gap will close: “Building agentic capabilities in base LLMs has become a key focus for many LLM builders, and once people realize the value of this, it will only be a matter of time." For now, Manus is doubling down on audiences it can already serve. In a written response, the company said its “primary focus is overseas expansion,” noting that new offices in San Francisco, Singapore, and Tokyo have opened in the past month. A super‑app approach Although the concept of AI agents is still relatively new, the consumer-facing AI app market in China is already crowded with major tech players. DeepSeek remains the most widely used, while ByteDance’s Doubao and Moonshot’s Kimi have also become household names. However, most of these apps are still optimized for chat and entertainment rather than task execution. This gap in the local market has pushed China’s big tech firms to roll out their own user-facing agents, though early versions remain uneven in quality and rough around the edges.  ByteDance is testing Coze Space, an AI agent based on its own Doubao model family that lets users toggle between “plan” and “execute” modes, so they can either directly guide the agent’s actions or step back and watch it work autonomously. It connects up to 14 popular apps, including GitHub, Notion, and the company’s own Lark office suite. Early reviews say the tool can feel clunky and has a high failure rate, but it clearly aims to match what Manus offers. Meanwhile, Zhipu AI has released a free agent called AutoGLM Rumination, built on its proprietary ChatGLM models. Shanghai‑based Minimax has launched Minimax Agent. Both products look almost identical to Manus and demo basic tasks such as building a simple website, planning a trip, making a small Flash game, or running quick data analysis. Despite the limited usability of most general AI agents launched within China, big companies have plans to change that. During a May 15 earnings call, Tencent president Liu Zhiping teased an agent that would weave automation directly into China’s most ubiquitous app, WeChat.  Considered the original super-app, WeChat already handles messaging, mobile payments, news, and millions of mini‑programs that act like embedded apps. These programs give Tencent, its developer, access to data from millions of services that pervade everyday life in China, an advantage most competitors can only envy. Historically, China’s consumer internet has splintered into competing walled gardens—share a Taobao link in WeChat and it resolves as plaintext, not a preview card. Unlike the more interoperable Western internet, China’s tech giants have long resisted integration with one another, choosing to wage platform war at the expense of a seamless user experience. But the use of mini‑programs has given WeChat unprecedented reach across services that once resisted interoperability, from gym bookings to grocery orders. An agent able to roam that ecosystem could bypass the integration headaches dogging independent startups. Alibaba, the e-commerce giant behind the Qwen model series, has been a front-runner in China's AI race but has been slower to release consumer-facing products. Even though Qwen was the most downloaded open-source model on Hugging Face in 2024, it didn’t power a dedicated chatbot app until early 2025. In March, Alibaba rebranded its cloud storage and search app Quark into an all-in-one AI search tool. By June, Quark had introduced DeepResearch—a new mode that marks its most agent-like effort to date.  ByteDance and Alibaba did not reply to MIT Technology Review’s request for comments. “Historically, Chinese tech products tend to pursue the all-in-one, super-app approach, and the latest Chinese AI agents reflect just that,” says Li of Simular, who previously worked at Google DeepMind on AI-enabled work automation. “In contrast, AI agents in the US are more focused on serving specific verticals.” Pei, the researcher at Stanford, says that existing tech giants could have a huge advantage in bringing the vision of general AI agents to life—especially those with built-in integration across services. “The customer-facing AI agent market is still very early, with tons of problems like authentication and liability,” he says. “But companies that already operate across a wide range of services have a natural advantage in deploying agents at scale.”

Join the event trusted by enterprise leaders for nearly two decades. VB Transform brings together the people building real enterprise AI strategy. Learn more Google has released an updated preview of​​ Gemini 2.5 Pro, its “most intelligent” model, first announced in March and upgraded in May, as a preview, intending to release the same model to general availability in a couple of weeks.  Enterprises can test building new applications or replace earlier versions with an updated version of the “I/O edition” of Gemini 2.5 Pro that, according to a blog post by Google, is more creative in its responses and outperforms other models in coding and reasoning.  During its annual I/O developer conference in May, Google announced that it updated Gemini 2.5 Pro to be better than its earlier iteration, which it quietly released. Google DeepMind CEO Demis Hassabis said the I/O edition is the company’s best coding model yet.  But this new preview, called Gemini 2.5 Pro Preview 06-05 Thinking, is even better than the I/O edition. The stable version Google plans to release publicly is “ready for enterprise-scale capabilities.” The I/O edition, or gemini-2.5-pro-preview-05-06, was first made available to developers and enterprises in May through Google AI Studio and Vertex AI. Gemini 2.5 Pro Preview 06-05 Thinking can be accessed via the same platforms.  Performance metrics This new version of Gemini 2.5 Pro performs even better than the first release.  Google said the new version of Gemini 2.5 Pro improved by 24 points in LMArena and by 35 points in WebDevArena, where it currently tops the leaderboard. The company’s benchmark tests showed that the model outscored competitors like OpenAI’s o3, o3-mini, and o4-mini, Anthropic’s Claude 4 Opus, Grok 3 Beta from xAI and DeepSeek R1.  “We’ve also addressed feedback from our previous 2.5 Pro releases, improving its style and structure — it can be more creative with better-formatted responses,” Google said in the blog post.  What enterprises can expect Google’s continuous improvement of Gemini 2.5 Pro might be confusing for many, but Google previously framed these as a response to community feedback. Pricing for the new version is $1.25 per million tokens without caching for inputs and $10 for the output price.  When the very first version of Gemini 2.5 Pro launched in March, VentureBeat’s Matt Marshall called it “the smartest model you’re not using.” Since then, Google has integrated the model into many of its new applications and services, including “Deep Think,” where Gemini considers multiple hypotheses before responding.  The release of Gemini 2.5 Pro, and its two upgraded versions, revived Google’s place in the large language model space after competitors like DeepSeek and OpenAI diverted the industry’s attention to their reasoning models.  In just a few hours of announcing the updated Gemini 2.5 Pro, developers have already begun playing around with it. While many found the update to live up to Google’s promise of being faster, the jury is still out if this latest Gemini 2.5 Pro does actually perform better.  Daily insights on business use cases with VB Daily If you want to impress your boss, VB Daily has you covered. We give you the inside scoop on what companies are doing with generative AI, from regulatory shifts to practical deployments, so you can share insights for maximum ROI. Read our Privacy Policy Thanks for subscribing. Check out more VB newsletters here. An error occured.

Last year, China saw a boom in foundation models, the do-everything large language models that underpin the AI revolution. This year, the focus has shifted to AI agents—systems that are less about responding to users’ queries and more about autonomously accomplishing things for them.  There are now a host of Chinese startups building these general-purpose digital tools, which can answer emails, browse the internet to plan vacations, and even design an interactive website. Many of these have emerged in just the last two months, following in the footsteps of Manus—a general AI agent that sparked weeks of social media frenzy for invite codes after its limited-release launch in early March.  These emerging AI agents aren’t large language models themselves. Instead, they’re built on top of them, using a workflow-based structure designed to get things done. A lot of these systems also introduce a different way of interacting with AI. Rather than just chatting back and forth with users, they are optimized for managing and executing multistep tasks—booking flights, managing schedules, conducting research—by using external tools and remembering instructions.  China could take the lead on building these kinds of agents. The country’s tightly integrated app ecosystems, rapid product cycles, and digitally fluent user base could provide a favorable environment for embedding AI into daily life.  For now, its leading AI agent startups are focusing their attention on the global market, because the best Western models don’t operate inside China’s firewalls. But that could change soon: Tech giants like ByteDance and Tencent are preparing their own AI agents that could bake automation directly into their native super-apps, pulling data from their vast ecosystem of programs that dominate many aspects of daily life in the country.  As the race to define what a useful AI agent looks like unfolds, a mix of ambitious startups and entrenched tech giants are now testing how these tools might actually work in practice—and for whom. Set the standard It’s been a whirlwind few months for Manus, which was developed by the Wuhan-based startup Butterfly Effect. The company raised $75 million in a funding round led by the US venture capital firm Benchmark, took the product on an ambitious global roadshow, and hired dozens of new employees.  Even before registration opened to the public in May, Manus had become a reference point for what a broad, consumer‑oriented AI agent should accomplish. Rather than handling narrow chores for businesses, this “general” agent is designed to be able to help with everyday tasks like trip planning, stock comparison, or your kid’s school project.  Unlike previous AI agents, Manus uses a browser-based sandbox that lets users supervise the agent like an intern, watching in real time as it scrolls through web pages, reads articles, or codes actions. It also proactively asks clarifying questions, supports long-term memory that would serve as context for future tasks. “Manus represents a promising product experience for AI agents,” says Ang Li, cofounder and CEO of Simular, a startup based in Palo Alto, California, that’s building computer use agents, AI agents that control a virtual computer. “I believe Chinese startups have a huge advantage when it comes to designing consumer products, thanks to cutthroat domestic competition that leads to fast execution and greater attention to product details.” In the case of Manus, the competition is moving fast. Two of the most buzzy follow‑ups, Genspark and Flowith, for example, are already boasting benchmark scores that match or edge past Manus’s.  Genspark, led by former Baidu executives Eric Jing and Kay Zhu, links many small “super agents” through what it calls multi‑component prompting. The agent can switch among several large language models, accepts both images and text, and carries out tasks from making slide decks to placing phone calls. Whereas Manus relies heavily on Browser Use, a popular open-source product that lets agents operate a web browser in a virtual window like a human, Genspark directly integrates with a wide array of tools and APIs. Launched in April, the company says that it already has over 5 million users and over $36 million in yearly revenue. Flowith, the work of a young team that first grabbed public attention in April 2025 at a developer event hosted by the popular social media app Xiaohongshu, takes a different tack. Marketed as an “infinite agent,” it opens on a blank canvas where each question becomes a node on a branching map. Users can backtrack, take new branches, and store results in personal or sharable “knowledge gardens”—a design that feels more like project management software (think Notion) than a typical chat interface. Every inquiry or task builds its own mind-map-like graph, encouraging a more nonlinear and creative interaction with AI. Flowith’s core agent, NEO, runs in the cloud and can perform scheduled tasks like sending emails and compiling files. The founders want the app to be a “knowledge marketbase”, and aims to tap into the social aspect of AI with the aspiration of becoming “the OnlyFans of AI knowledge creators”. What they also share with Manus is the global ambition. Both Genspark and Flowith have stated that their primary focus is the international market. A global address Startups like Manus, Genspark, and Flowith—though founded by Chinese entrepreneurs—could blend seamlessly into the global tech scene and compete effectively abroad. Founders, investors, and analysts that MIT Technology Review has spoken to believe Chinese companies are moving fast, executing well, and quickly coming up with new products.  Money reinforces the pull to launch overseas. Customers there pay more, and there are plenty to go around. “You can price in USD, and with the exchange rate that’s a sevenfold multiplier,” Manus cofounder Xiao Hong quipped on a podcast. “Even if we’re only operating at 10% power because of cultural differences overseas, we’ll still make more than in China.” But creating the same functionality in China is a challenge. Major US AI companies including OpenAI and Anthropic have opted out of mainland China because of geopolitical risks and challenges with regulatory compliance. Their absence initially created a black market as users resorted to VPNs and third-party mirrors to access tools like ChatGPT and Claude. That vacuum has since been filled by a new wave of Chinese chatbots—DeepSeek, Doubao, Kimi—but the appetite for foreign models hasn’t gone away.  Manus, for example, uses Anthropic’s Claude Sonnet—widely considered the top model for agentic tasks. Manus cofounder Zhang Tao has repeatedly praised Claude’s ability to juggle tools, remember contexts, and hold multi‑round conversations—all crucial for turning chatty software into an effective executive assistant. But the company’s use of Sonnet has made its agent functionally unusable inside China without a VPN. If you open Manus from a mainland IP address, you’ll see a notice explaining that the team is “working on integrating Qwen’s model,” a special local version that is built on top of Alibaba’s open-source model.  An engineer overseeing ByteDance’s work on developing an agent, who spoke to MIT Technology Review anonymously to avoid sanction, said that the absence of Claude Sonnet models “limits everything we do in China.” DeepSeek’s open models, he added, still hallucinate too often and lack training on real‑world workflows. Developers we spoke with rank Alibaba’s Qwen series as the best domestic alternative, yet most say that switching to Qwen knocks performance down a notch. Jiaxin Pei, a postdoctoral researcher at Stanford’s Institute for Human‑Centered AI, thinks that gap will close: “Building agentic capabilities in base LLMs has become a key focus for many LLM builders, and once people realize the value of this, it will only be a matter of time.” For now, Manus is doubling down on audiences it can already serve. In a written response, the company said its “primary focus is overseas expansion,” noting that new offices in San Francisco, Singapore, and Tokyo have opened in the past month. A super‑app approach Although the concept of AI agents is still relatively new, the consumer-facing AI app market in China is already crowded with major tech players. DeepSeek remains the most widely used, while ByteDance’s Doubao and Moonshot’s Kimi have also become household names. However, most of these apps are still optimized for chat and entertainment rather than task execution. This gap in the local market has pushed China’s big tech firms to roll out their own user-facing agents, though early versions remain uneven in quality and rough around the edges.  ByteDance is testing Coze Space, an AI agent based on its own Doubao model family that lets users toggle between “plan” and “execute” modes, so they can either directly guide the agent’s actions or step back and watch it work autonomously. It connects up to 14 popular apps, including GitHub, Notion, and the company’s own Lark office suite. Early reviews say the tool can feel clunky and has a high failure rate, but it clearly aims to match what Manus offers. Meanwhile, Zhipu AI has released a free agent called AutoGLM Rumination, built on its proprietary ChatGLM models. Shanghai‑based Minimax has launched Minimax Agent. Both products look almost identical to Manus and demo basic tasks such as building a simple website, planning a trip, making a small Flash game, or running quick data analysis. Despite the limited usability of most general AI agents launched within China, big companies have plans to change that. During a May 15 earnings call, Tencent president Liu Zhiping teased an agent that would weave automation directly into China’s most ubiquitous app, WeChat.  Considered the original super-app, WeChat already handles messaging, mobile payments, news, and millions of mini‑programs that act like embedded apps. These programs give Tencent, its developer, access to data from millions of services that pervade everyday life in China, an advantage most competitors can only envy. Historically, China’s consumer internet has splintered into competing walled gardens—share a Taobao link in WeChat and it resolves as plaintext, not a preview card. Unlike the more interoperable Western internet, China’s tech giants have long resisted integration with one another, choosing to wage platform war at the expense of a seamless user experience. But the use of mini‑programs has given WeChat unprecedented reach across services that once resisted interoperability, from gym bookings to grocery orders. An agent able to roam that ecosystem could bypass the integration headaches dogging independent startups. Alibaba, the e-commerce giant behind the Qwen model series, has been a front-runner in China’s AI race but has been slower to release consumer-facing products. Even though Qwen was the most downloaded open-source model on Hugging Face in 2024, it didn’t power a dedicated chatbot app until early 2025. In March, Alibaba rebranded its cloud storage and search app Quark into an all-in-one AI search tool. By June, Quark had introduced DeepResearch—a new mode that marks its most agent-like effort to date.  ByteDance and Alibaba did not reply to MIT Technology Review’s request for comments. “Historically, Chinese tech products tend to pursue the all-in-one, super-app approach, and the latest Chinese AI agents reflect just that,” says Li of Simular, who previously worked at Google DeepMind on AI-enabled work automation. “In contrast, AI agents in the US are more focused on serving specific verticals.” Pei, the researcher at Stanford, says that existing tech giants could have a huge advantage in bringing the vision of general AI agents to life—especially those with built-in integration across services. “The customer-facing AI agent market is still very early, with tons of problems like authentication and liability,” he says. “But companies that already operate across a wide range of services have a natural advantage in deploying agents at scale.”

Interested in trying out Claude Code? Now you can, for less.

Interested in trying out Claude Code? Now you can, for less.

Anthropic says that it has released a new set of AI models tailored for U.S. national security customers. The new models, a custom set of “Claude Gov” models, were “built based on direct feedback from our government customers to add

Anthropic’s AI-written blog Claude Explains works best when it sticks to writing about itself and uses human editing.