Anthropic, backed by Amazon, announces breakthrough as AI agents achieve human-like task completion

Anthropic, the Amazon-backed AI startup founded by former OpenAI executives, announced on Tuesday it has reached a significant milestone in artificial intelligence: AI agents capable of using computers to complete complex tasks in a similar way to humans.

The startup is behind Claude, one of the chatbots competing with OpenAI’s ChatGPT and Google’s Gemini, which have all surged in popularity. Companies like Anthropic, alongside tech giants such as Google, Amazon, Microsoft, and Meta, are part of a rapidly intensifying race in generative AI, a market forecast to exceed $1 trillion in revenue within the next decade.

Anthropic’s breakthrough, dubbed "Computer Use," is part of two newly introduced AI models. This technology allows its AI to interpret what’s on a computer screen, select buttons, enter text, browse the internet, and perform tasks across different software platforms.

Speaking to CNBC, Jared Kaplan, Anthropic's Chief Science Officer, explained the tool's functionality: “It can use computers in basically the same way that we do,” adding that the system can manage “tens or even hundreds of steps” to complete tasks.

The tool was initially made available to early access partners like Amazon and companies such as Asana, Canva, and Notion. Anthropic has been developing this capability since the start of the year and released it to developers in public beta on Tuesday. The startup hopes to extend access to consumers and businesses over the coming months or by early next year, Kaplan said.

Potential consumer uses include booking flights, scheduling appointments, completing forms, conducting research online, and filing expense reports. “We want Claude to assist people with all sorts of work,” Kaplan said, adding that while chatbots answer questions, “the current setup is limited” since it doesn’t allow further context or action beyond that.

AI agents, like the ones being developed by Anthropic, represent a step beyond chatbots by being capable of handling more complex, multi-step tasks. Unlike traditional chatbots that simply respond to queries, these agents are designed to execute tasks on behalf of users, potentially transforming business operations. Kaplan described these agents as evolving from virtual assistants to "virtual collaborators."

The push toward AI agents has gained significant traction across the tech world. Microsoft CEO Satya Nadella has said that his company is also working towards creating AI agents to help users complete more tasks, though he acknowledged the challenges ahead. Executives from Meta and Google have similarly voiced their ambitions to make AI agents more productive.

Since its founding, Anthropic has become a key player in the AI landscape. The company first launched its Claude chatbot in March 2023, positioning itself as a direct competitor to OpenAI’s ChatGPT in both business and consumer markets. Despite limited consumer availability, Anthropic’s backers include major investors such as Google, Salesforce, and Amazon.

In recent months, Anthropic has introduced iOS and Android apps, as well as a “Team” plan for businesses, expanding its presence in Europe. Its latest product, Claude Enterprise, is aimed at companies seeking to integrate AI into their operations. Early users of Claude Enterprise include GitLab, Midjourney, and Menlo Ventures.

The Claude Enterprise product offers customers an expanded context window for data uploads—allowing businesses to process the equivalent of 100 30-minute sales calls, 100,000 lines of code, or 15 full financial reports at once. It also includes “activity feeds” that allow AI veterans within companies to show new users how they are employing the technology.

Anthropic’s progress in AI didn’t stop there. In June, it introduced "Artifacts," or "workspaces," where users can generate text documents, code, and edit these in real time with Claude’s assistance. The system enables enterprise clients to create marketing plans, process sales data, draft legal documents, and more.

The company has also added key industry figures to its team. Mike Krieger, co-founder and former chief technology officer of Instagram, joined as Chief Product Officer in May. Under Krieger’s leadership, Instagram grew to 1 billion users, and its engineering team expanded to over 450 members. Jan Leike, a former safety leader at OpenAI, also joined Anthropic that same month.

Anthropic’s recent announcements reflect its long-term ambition to build AI agents that operate as more than just assistants—evolving into highly capable virtual collaborators in business and beyond.