Smartsheet is preparing a fundamental shift in how work management platforms interact with artificial intelligence. In June 2026, the enterprise-grade platform will roll out a Model Context Protocol (MCP) Server that natively connects with ChatGPT, Microsoft Copilot, and Google Cloud Gemini Enterprise, while simultaneously introducing Smart Assist—a new AI-powered capability embedded directly inside the Smartsheet experience. The move aims to turn Smartsheet from a passive work repository into an active, AI-driven orchestration layer for complex projects and portfolios.

MCP, originally developed by Anthropic as an open standard, allows large language models to securely access external data sources and tools without custom integrations. Smartsheet’s adoption of the protocol signals a deliberate bet on interoperability at a time when enterprises struggle with fragmented AI assistants scattered across email, documents, and communication channels. By standing up an MCP Server, Smartsheet ensures that any MCP-compatible client—whether it’s ChatGPT, Copilot, or Gemini—can query, update, and reason over Smartsheet data in a governed, auditable way.

What the MCP Server Actually Delivers

For the tens of thousands of organizations managing projects, resources, and workflows inside Smartsheet, the MCP Server turns static sheets, reports, and dashboards into live AI-accessible resources. A user inside ChatGPT could ask, “What is the status of the Q3 product launch, and who is blocking the critical path?” The AI would call the Smartsheet MCP Server, retrieve the relevant sheet data, compute dependencies, and return a human-readable answer—without ever leaving the chat interface.

Microsoft Copilot will benefit from similar deep linking. Inside Teams or Word, a project manager could instruct Copilot to pull a risk register directly from Smartsheet and summarize it in an executive memo. Google Cloud Gemini Enterprise users, meanwhile, gain the ability to incorporate Smartsheet context into BigQuery analyses or Vertex AI pipelines, bridging structured work data with advanced analytics.

Crucially, the MCP Server is designed to act as a trust boundary. Administrators will set granular permission scopes that determine which AI assistants can see which workspaces, sheets, or specific fields. Every access request flows through Smartsheet’s existing governance model, so row-level sharing rules, cell-level permissions, and audit logs remain intact. This addresses the top concern IT leaders voice about connecting AI copilots to enterprise data: loss of control over who sees what.

Introducing Smart Assist inside Smartsheet

Alongside the external connections, Smartsheet is building Smart Assist—a native AI interface embedded directly into the Smartsheet web and mobile apps. Unlike the remote AI assistants that call the MCP Server from outside, Smart Assist lives inside the platform and understands the full context of the user’s current view.

Early demonstrations show Smart Assist capable of generating complex formulas from natural language descriptions, drafting project summary narratives based on live sheet data, and suggesting risk mitigation actions when deadlines slip. A portfolio manager staring at a dashboard of 40 active projects can simply type, “Which projects are more than 20% over budget and what are the common causes?” Smart Assist scans the underlying sheets, correlates financial data, and surfaces patterns—no manual filtering required.

Power users will appreciate Smart Assist’s ability to automate routine housekeeping. It can be instructed to watch for newly created rows lacking assigned owners, then to reach out via Slack or email to the nearest manager. It can generate weekly status reports that blend quantitative metrics with AI-composed commentary, pulling data from cross-sheet references and attachment metadata.

How Smartsheet MCP Differs from Standard API Connectors

The shift from proprietary APIs to the MCP standard is more than a technical convenience. APIs force each AI product to build and maintain a custom Smartsheet integration, which historically led to inconsistent capabilities and delayed updates. By contrast, the MCP Server exposes a uniform tool set—actions like read-sheet, update-cell, and execute-report—that any compliant AI assistant can invoke immediately. This means enterprises get consistent behavior whether a user is interacting through ChatGPT’s web interface, Copilot’s side panel in Excel, or a Gemini-powered internal chatbot.

Smartsheet’s decision to support MCP also helps future-proof its AI strategy. As newer reasoning models or specialized finance AIs emerge, they too can access Smartsheet data the day they support MCP, with no additional work from Smartsheet engineering teams. For customers burned by vendor lock-in on previous workflow automation platforms, this open-ecosystem approach reduces the risk of being stranded with an AI integration that only works with one vendor’s particular model.

Enterprise AI Governance Implications

Connecting Smartsheet to the three largest enterprise AI platforms raises thorny governance questions that Smartsheet appears to have anticipated. The MCP Server design includes an intermediary layer that sanitizes outgoing data—users can configure it to redact personally identifiable information, mask financial figures below a certain threshold, or strip internal comments before they reach the external AI. This ensures that even when a user inadvertently asks an unsecured AI instance about a sensitive sheet, the MCP Server applies policy before any data leaves Smartsheet’s boundary.

Every MCP call produces an immutable audit event, logged alongside other Smartsheet activity. Compliance officers can trace exactly which AI, prompted by which user, accessed which cell at what time. Such granularity will be mandatory for organizations subject to SOX, HIPAA, or GDPR, where demonstrating control over AI data flows is rapidly becoming a regulatory expectation.

Smart Assist operates under the same governance framework. When a user asks Smart Assist to summarize a sheet that contains confidential HR data, the AI will only use fields the user is permitted to see. If the request crosses into territory that requires sign-off—such as sending automated communications to external parties—Smart Assist can enforce a human-approval step before executing.

Real-World Impact on Work Management

The combination of external AI connectivity and an internal Smart Assist effectively creates a dual-layer AI service. The external layer extends Smartsheet’s data into the tools knowledge workers already live in—chat apps, office suites, analytics platforms—while the internal layer optimizes the dedicated Smartsheet experience for power users who spend much of their day managing complex plans.

Consider a construction project office: A field supervisor updates a daily progress log in Smartsheet from a tablet. As soon as the row is saved, Smart Assist detects that weather-related delays are accumulating and automatically adjusts downstream task start dates across five linked sheets. Simultaneously, an executive using Copilot in Teams asks for a program-level view of delays across all active projects; Copilot queries the Smartsheet MCP Server and returns a consolidated risk report in seconds, complete with a recommended action plan sourced from Smartsheet’s historical project data. No human manually compiled that report.

This fluidity addresses a longstanding pain point in work management software: the disconnect between transactional data entry and strategic decision-making. By making Smartsheet data AI-accessible, decisions no longer depend on someone proactively building a dashboard or exporting a CSV. The AI can pull the data at the moment of need, in context, and transform it into insight.

Pricing, Availability, and Early Feedback

Smartsheet has confirmed that the MCP Server and Smart Assist will launch in June 2026 as part of the Advance and Enterprise plan tiers. Specific pricing details have not been disclosed, but company executives indicated during a pre-briefing that they are “committed to making AI a core capability, not a premium add-on that breaks budgets.” Analysts interpret this to mean the features will be included in existing plan fees, at least initially, with possible consumption-based charges for high-volume AI interactions introduced later.

Early-access test participants report significant time savings—one program manager at a multinational pharmaceutical company noted that Smart Assist reduced the time spent compiling monthly portfolio reviews from twelve hours to under two hours, largely by automating data aggregation and narrative generation. Others praised the MCP Server’s ability to expose Smartsheet data to ChatGPT, but cautioned that the quality of AI output depends heavily on the clarity of the underlying sheet structure. Poorly designed sheets with inconsistent column naming produce less reliable AI results, a limitation Smartsheet acknowledges and is addressing through enhanced templates and AI-powered data normalization features scheduled to arrive before general availability.

The Competitive Landscape

Smartsheet’s move intensifies competition with the Asana-Monday.com-Notion trio, all of whom have invested heavily in native AI assistants but have not yet embraced cross-platform protocol standards like MCP. Asana’s AI features remain largely confined to its own interface, while Monday.com’s AI is tightly integrated with its work operating system but lacks native MCP support. Notion AI is powerful within the Notion universe but cannot serve as a data source for external copilots without custom API work.

By adopting MCP, Smartsheet positions itself as the neutral work management layer that feeds any AI assistant. This could become a decisive advantage in RFPs where CIOs demand that project data be accessible across their entire AI toolchain, not siloed inside whichever platform first shipped a chatbot.

Microsoft’s own Project and Planner tools compete directly with Smartsheet, and while they natively integrate with Copilot, they do not offer the same level of platform agnosticism. A Copilot user can manage tasks inside Microsoft 365, but if the same user also relies on ChatGPT for high-level strategy work or Gemini for data science, only Smartsheet can seamlessly feed both assistants the same up-to-date project data.

Looking Ahead: AI-First Work Management

The June 2026 launch is not an endpoint but a foundation. Smartsheet’s product roadmap, shared under NDA with key customers, points to MCP Server extensions that will eventually expose workflow approvals, resource management tools, and even custom Smartsheet automations as AI-invocable tools. If a VP of Engineering can ask Gemini, “Re-allocate the overbooked electrical engineers to reduce project delay risk,” and the AI—with appropriate permissions—actually reshuffles assignments across Smartsheet sheets, that would represent a paradigm shift in how work gets managed.

Smart Assist is also set to evolve. Future iterations will leverage multimodal models capable of processing screenshots of hand-drawn Gantt charts, interpreting voice-annotated task updates, and even proactively scheduling check-ins when it detects potential bottlenecks. Smartsheet is exploring a “copilot-as-advisor” mode where the AI runs continuous Monte Carlo simulations on active project data and privately notifies managers when risk thresholds are breached, long before a status meeting would reveal the problem.

Practical Steps for Smartsheet Admins

Organizations planning to adopt the MCP Server should begin preparing now. Smartsheet recommends auditing existing sheet structures for consistency, establishing a governance working group that includes IT, compliance, and business unit leads, and running pilot programs with synthetic data to understand how AI outputs align with business expectations. Cleansing legacy sheets—renaming columns consistently, eliminating duplicate fields, and documenting key business logic—will dramatically improve the quality of AI interactions once the MCP Server is live.

Training materials and a sandbox environment are expected to be made available to current Advance and Enterprise customers by April 2026, giving teams two months to test integrations before the full launch. Early adopters are urged to engage with Smartsheet’s professional services team to design permission models that balance AI utility with data security.

Smartsheet’s June 2026 announcement is more than a feature update. It signals a future where the line between project management software and AI orchestration fades, and where the value of a work management platform is measured not just by how well it organizes tasks, but by how effectively it feeds the AI assistants that increasingly make decisions.