anthropic announced on Monday that users can now open and interact with popular business applications directly inside cloudThe company’s AI assistant—a key extension that transforms the chatbot from a conversational tool into an integrated workspace where employees can create project timelines, draft Slack messages, create presentations, and visualize data without having to switch browser tabs.
The rollout, which goes live today, includes integration Dimensions, Posture, box, Canva, Soil, figma, hex, somwar.comAnd Loose. Salesforce integration coming soon. The feature marks a new chapter in Anthropic’s aggressive effort to dominate enterprise AI, coming just days after the company’s CEO made headlines at Davos with bold predictions about AI replacing white-collar workers.
"MCP apps are extensions of the core MCP protocol and are part of the open source MCP ecosystem," Sean Strong, Anthropic’s product manager for MCP apps, told VentureBeat in an exclusive interview. "Within Claude.ai, connectors require a paid cloud plan – Pro, Max, Team, or Enterprise – but there are no additional fees associated with using connectors."
That pricing decision is noteworthy. Rather than monetizing integrations separately or charging partners for delivery, Anthropic is bundling interactive tools into existing subscription tiers — a strategy designed to accelerate adoption and deepen cloud penetration in corporate environments where the company reportedly already leads OpenAI.
Inside MCP Apps, the open-source technology that lets the cloud control your favorite work tools
The technical basis is what is called anthropic "mcp apps," A new extension of the Model Context Protocol, an open standard for connecting external devices to AI applications that Anthropic open-sourced last year. mcp apps Allow any MCP server to deliver an interactive user interface within any assistive AI product – meaning the technology is not limited to the cloud.
In practice, integration allows surprisingly granular control. Users can create analytics charts in Dimensions And adjust parameters interactively to detect trends. They can turn conversations into Asana projects with tasks and deadlines that sync automatically. They can inspire Cloud to create a flowchart or Gantt chart in Figma’s collaborative whiteboard tool, Figjam. They can draft Slack messages, preview the formatting, and review before posting.
Hex integration may prove particularly valuable for data teams: Users can ask data questions in natural language and receive full answers with interactive charts, tables, and quotes – effectively turning the cloud into a business intelligence interface.
"We have open sourced MCP to give the ecosystem a universal way to connect devices to AI," the company said in its announcement blog. "We’re now further extending MCP so that developers can build interactive UIs on top of it, no matter where their users are."
What happens when AI can send messages and create projects on your behalf
As AI systems become increasingly capable of taking real-world actions – sending messages, creating projects, publishing content – the question of guardrails becomes important. Could an employee accidentally send an unreviewed Slack message or publish an incomplete Canva presentation?
Strong addressed this directly. "Most major MCP clients, including cloud ones, provide consent prompts that help users determine whether they want to perform any action through the MCP server," He said.
For enterprise deployments, IT administrators maintain control. "Team and enterprise admins have the ability to control which MCP servers users in their organization have the ability to access," Explained strongly.
The consent-signal approach is a middle ground between complete autonomy and a cumbersome approval workflow. But it also places significant responsibility on individual users to review actions before confirming them – a design choice that may attract scrutiny as AI agents become capable of making more consequential decisions.
Security concerns are not imaginary. As Fortune reported last week, Anthropic’s Cloud Code product faces vulnerabilities "quick injection," Where attackers hide malicious instructions in web content to manipulate AI behavior. The company has implemented several security layers, including running certain features in virtual machines and adding deletion protection after users accidentally delete files. "Agent security – that is, the act of securing the real-world activities of the cloud – is still an active area of development in the industry," Anthropic has accepted.
Cloud Code’s viral success sets the stage for Anthropic’s enterprise ambitions
The announcement of the interactive tool comes at a moment of unusual momentum for Anthropic. The company’s coding assistant Cloud Code, released in February 2024, has become a viral hit that has attracted attention far beyond its intended developer audience.
Originally built for software developers, Cloud Code has attracted attention far beyond its intended audience. Non-programmers have deployed it to book theater tickets, file taxes, and monitor tomato plants. Nvidia CEO Jensen Huang called it "Incredible." Even Microsoft, which sells competitor GitHub Copilot, has widely adopted Cloud Code internally, reportedly encouraging non-developers to use it.
Boris Cherny, Anthropic’s head of cloud code, told Fortune that his team built Cowork — a user-friendly version of the coding product for non-programmers — in about a week and a half, using largely cloud code. "Engineers now feel liberated, that they no longer have to work on all the difficult things," Cherney said.
According to Anthropic, Cloud Code is now used by Uber, Netflix, Spotify, Salesforce, Accenture, and Snowflake. According to data from SimilarWeb and Sensor Tower published by The Wall Street Journal, the cloud’s total web audience has more than doubled by December 2024, and daily unique visitors on desktop are up 12% year-over-year.
The company is also reportedly planning a $10 billion fundraising round, which would value Anthropic at $350 billion – a staggering figure that reflects investors’ confidence in the company’s venture attractiveness.
Anthropic CEO causes controversy at Davos with predictions about AI replacing workers
The launch of the interactive tool also comes against the backdrop of an intense debate about the impact of AI on employment – a debate that Anthropic’s own CEO helped sharpen at the World Economic Forum in Davos last week.
Dario Amodei tells Davos audience that AI models will replace the work of all software developers within a year and will reach "Nobel level" Scientific research in multiple fields within two years. He predicted that 50% of white-collar jobs would disappear within five years.
"I have engineers at Anthropic who say ‘I don’t write any code anymore. I just let the model write the code, I edit it,’" Amodei said. "We may be six to 12 months away from the time when the model is doing most of the work, probably all that the software engineers do from start to finish."
Not everyone agrees with that timeline. Demis Hassabis, the Nobel Prize-winning CEO of Google DeepMind, said at the same conference that today’s AI systems "nowhere near" Human-level artificial general intelligence. Yann LeCun, the Turing Award-winning AI pioneer who recently left META to found Advanced Machine Intelligence Labs, went further, arguing that large language models "will never be able to attain human intelligence" And a completely different approach is required.
Why embedding AI into daily workflows can create a powerful lock-in for enterprises?
Anthropic’s integration strategy reflects broader changes in the enterprise AI competition. The battlefield is moving from model benchmarks and capability demonstrations to workflow integration – the extent to which AI systems become embedded in how companies actually operate.
By making the cloud the interface through which employees interact with Asana, Slack, Figma, and other daily tools, Anthropic is positioning itself not just as an AI provider but as a workflow orchestration layer. The more operations flow through the cloud, the more difficult it becomes for enterprises to switch to a competitor.
This approach mirrors strategies that proved successful for previous generations of enterprise software. Salesforce built its dominance in part by becoming the system of record for customer data. Centralizing workplace communications increased tardiness. Anthropic appears to be betting that AI assistants can occupy a similar position – the default starting point for work.
MCP’s open-source foundation can accelerate this strategy. By making the protocol available to any developer, Anthropic encourages a broad ecosystem of integration that funnels through all MCP-compliant clients – of which the cloud is the foremost. The company benefits from network effects even if it keeps the standard open.
The race to become the operating system for AI-powered work is just beginning
The launch notably excludes some major enterprise platforms. The salesforce integration is listed as "coming soon," And that’s with no mention of Microsoft 365, Google Workspace, or other productivity suites that dominate corporate environments. Those gaps may limit early adoption in organizations heavily invested in those ecosystems.
The feature is available on the web and desktop for paid cloud plans, with support for Cloud Cowork – the file management agent launched last week – coming later. Mobile support was not mentioned in the announcement.
For enterprises evaluating OpenAI offerings and the cloud against other competitors, interactive integrations represent a solid differentiator. The ability to take actions within business tools – rather than simply generating text that users must copy elsewhere – addresses a persistent friction point in AI adoption.
Whether that advantage proves sustainable depends on how quickly competitors respond. OpenAI has its own enterprise ambitions and partnerships. Google is integrating Gemini into its productivity suite. Microsoft continues to deepen CoPilot’s presence in Office applications.
But the bigger significance may be that today’s announcement signals where enterprise software is headed. For decades, the default unit of work has been applications – spreadsheets, project trackers, messaging platforms. Anthropic is betting that the future belongs to an AI layer that sits on top of them all.
If the company is right, the question for every enterprise software vendor becomes uncomfortably simple: Do you want to be the device, or the thing that controls the device?
<a href