Yazılar

Meta to buy Chinese-founded startup Manus to boost advanced AI

/in /tarafından

Meta said on Monday it will acquire Chinese-founded artificial intelligence startup Manus, stepping up efforts to integrate more advanced AI capabilities across its platforms. Financial terms were not disclosed, but a source with direct knowledge of the matter said the deal values the Singapore-based firm at between $2 billion and $3 billion.

Manus did not immediately respond to a request for comment. The startup drew widespread attention earlier this year after releasing what it described as the world’s first general AI agent—software designed to make decisions and execute tasks autonomously with far less prompting than conventional chatbots such as ChatGPT or DeepSeek. The launch sparked viral discussion on X and led some commentators to label Manus “China’s next DeepSeek,” with praise from Chinese state television.

Months later, Manus moved its headquarters from China to Singapore, joining a broader wave of Chinese-founded tech firms seeking to reduce exposure to rising U.S.-China tensions. The company’s products are not available in China. Manus has claimed its AI agent outperforms OpenAI’s DeepResearch and maintains a strategic partnership with Alibaba to collaborate on AI models.

Meta said it will operate and commercialize the Manus service and integrate it into both consumer and business offerings, including Meta AI. The acquisition reflects intensifying competition among large technology companies racing to secure differentiated AI capabilities through deals and talent hires.

Earlier this year, Meta invested in Scale AI in a transaction valuing the startup at $29 billion and bringing in its CEO, Alexandr Wang. Manus, backed by parent company Beijing Butterfly Effect Technology, raised $75 million this year at a valuation of about $500 million, according to the source, confirming prior media reports. The funding round was led by Benchmark, with investors including HSG, ZhenFund and Tencent Holdings, PitchBook data showed.

Amazon Sues Perplexity Over AI Shopping Agent That Secretly Accessed Customer Accounts

/in /tarafından

Amazon has filed a lawsuit against Perplexity AI, accusing the fast-growing startup of illegally accessing Amazon customer accounts through its automated “agentic” shopping feature. The complaint, filed Tuesday in a U.S. District Court in California, claims Perplexity’s Comet browser and AI agent disguised automated activity as human browsing to place orders on behalf of users.

The dispute marks a major flashpoint in the debate over regulating AI “agents” — autonomous digital assistants that can navigate websites, make purchases, and perform other online tasks on users’ behalf. Amazon said Perplexity’s technology posed security risks and had repeatedly ignored requests to stop unauthorized activity on its platform.

“Perplexity’s misconduct must end,” Amazon said in its filing, adding that the startup’s software “purposely disguised its automated activity” and that its actions were “no less unlawful” than a physical break-in.

Perplexity, whose AI tools have surged in popularity amid the rise of conversational assistants, previously dismissed Amazon’s complaints, calling them an attempt to stifle innovation and protect its ad-driven business model. “Bullying is when large corporations use legal threats to block innovation,” the startup said in an earlier blog post.

Amazon argued that Perplexity’s AI agent interfered with its ability to deliver a personalized shopping experience, undermining systems built over decades. The company added that third-party apps making purchases for users must act transparently and respect site policies.

Perplexity said its Comet AI assistant allows users to shop and compare products autonomously, while keeping login credentials stored locally, not on company servers. It argued that consumers have the right to choose their own AI shopping tools — a stance that could shape future legal battles over the limits of AI automation in e-commerce.

OpenAI set to launch AI-powered browser to rival Google Chrome

/in /tarafından

OpenAI is reportedly preparing to release a new AI-powered web browser in the coming weeks that aims to challenge Google Chrome’s dominance, according to sources familiar with the matter. The browser will leverage artificial intelligence to transform the way users interact with the web, potentially offering a more integrated experience that keeps some browsing activities within a ChatGPT-style chat interface rather than directing users to external websites.

This move marks a strategic push by OpenAI to gain direct access to user data—an essential asset for competing with Google, whose Chrome browser is a critical component of Alphabet’s advertising business, generating nearly 75% of its revenue. By controlling browsing data, OpenAI could directly rival Google in targeted advertising and user engagement.

With over 500 million weekly active ChatGPT users, OpenAI’s browser has the potential to significantly disrupt Google’s advertising ecosystem. The browser will be built on Chromium, the open-source codebase behind Chrome and other browsers like Microsoft Edge and Opera, allowing OpenAI to control data collection and integration more effectively.

The new browser is also designed to integrate OpenAI’s AI agent tools, such as Operator, enabling automated actions on behalf of users, including booking reservations or filling forms within websites—enhancing convenience and utility.

OpenAI’s founder Sam Altman, who has driven the company’s rapid innovation since ChatGPT’s launch in 2022, is betting on this browser as part of a broader strategy to embed AI deeper into daily personal and work life.

Competition is fierce: Google Chrome currently commands over two-thirds of the global browser market with more than 3 billion users, while Apple’s Safari holds a distant second place with 16%. Other AI-driven browsers like Comet by Perplexity, The Browser Company, and Brave have already launched AI-enhanced browsing experiences.

The Department of Justice’s ongoing antitrust actions against Google—following a ruling that Alphabet holds a monopoly in online search—highlight the regulatory challenges facing the search and browser giant. OpenAI has even expressed interest in acquiring Chrome if forced divestiture occurs.

Unlike merely creating a plug-in for existing browsers, OpenAI’s decision to build its own browser aims to maximize control over user data, a crucial factor for AI’s effectiveness and business value.

OpenAI declined to comment on the launch details.