Now released: New MCP Server Platform Activity Report

We have officially launched the new MCP Server Platform Activity report. This upcoming report will deliver broader platform-level metrics and visibility into overall MCP server environment activity, complementing the granular tool execution tracking found in the Tool Calls report. It includes detailed platform data to help administrators track and monitor what content in Box is being used through the MCP Server.

For details on what is tracked in this report, see our documentation here: https://docs.box.com/en/box-admin-tools/reporting-and-insights/mcp-platform-activity-report#mcp-platform-activity-report

Report Renaming: MCP Server Activity Report is now the "MCP Server Tool Calls Report"

To better reflect the specific insights it provides, the Box MCP Server Activity report is now renamed to the Box MCP Server Tool Calls report.

What This Report Covers

The Box MCP Server Tool Calls report provides an overview of tool usage executed through the Box MCP server. It includes detailed event-level data to help administrators track and audit MCP server interactions.

• Report Data Columns:
  • Date: The date (in Pacific Time, PT) of the MCP server activity.
  • Integration ID: The unique identifier of the integration used to access the MCP server.
  • Integration Name: The name of the integration used to access the MCP server.
  • Integration Client ID: The client ID associated with the integration.
  • User Name: The full name of the user who performed the activity.
  • User Email: The Box email address of the user who performed the activity.
• Available Filters:
  • Date Range (Optional): Limit the report to past MCP server activity within a specific time period.
  • Users or Groups (Optional): Filter activities by specific managed users or user groups (defaults to all users and groups).
  • Integrations (Optional): Filter by one or more integration names (defaults to all integrations).

What You Need to Do

No Action Required: Existing reporting configurations and scheduled workflows will automatically reflect the updated report name.

The latest version of the Box for Salesforce package is now available on Salesforce AppExchange. This release introduces new capabilities for Box Utility, along with improvements across Box for Salesforce features.

What's New

Box Utility – File Uploads

Box Utility now includes a File Uploads tab alongside Box Notes. From the Salesforce Utility Bar, users can upload files directly to the Box folder associated with a Salesforce record.

Users can upload either:

• Salesforce Files that are already attached to the record, or
• New files directly from their computer.

Uploaded files are saved to the same Box folder already linked to the Salesforce record. Learn more: Using Box Utility in Salesforce

Bug Fixes & Improvements

• Fixed an issue where Box Doc Gen generated documents with extra, incorrectly formatted tags for Salesforce fields that had no value.
• Fixed an issue where Box Doc Gen did not merge certain related field data when generating documents from records with multiple levels of related objects.
• Fixed the default minimum window size for Box Utility to ensure consistent display.

For questions or issues:

• Visit the Box Community
• Contact Box Support
• Review the Box for Salesforce documentation

We’ve released a new version of Box for Salesforce, now available for the download at the Salesforce AppExchange.

This release focuses on bug fixes and improvements to enhance the integration, including:

Box for Salesforce

Box Sign

• Improved Box Sign status sync consistency.
• Enhanced the Box Sign webhook to automatically retry when the Box Sign API status is behind the triggering event.
• Fixed an issue causing failed updates to appear as delivered skipping retry.
• Fixed an issue where the Box Sign scheduled sync reprocessed already-completed sign requests due to a truncated signer email value, causing repeated sync failures.
• Fixed an issue where a contact's phone number was not auto-populated when requesting SMS authentication for a sign request.
• Added a Box Sign Sync section to the Box Sign Configuration page in Box Settings, with a link to Using Box Sign Sync in Salesforce to help admins configure it correctly to receive real-time updates from Box Sign.

Box Doc Gen

• Fixed an issue where the Custom Object for Contact Details did not resolve correctly in Box Doc Gen templates.
• Fixed an issue where uploading a template in the Doc Gen Template Manager failed to collaborate the creating user on the new template folder.
• Fixed an issue where Box Doc Gen failed due to fields not present in the Salesforce API version.
• Added support for using lookups in Box Doc Gen name tags.

General

• Fixed an issue where collaborations on linked folders were removed or not added for the users.
• Added support for automatic collaboration in Box UI Elements when a Salesforce record ID is provided as input.
  • Note: Auto-collaboration does not apply when a hardcoded file or folder ID is used.
• Improvements to validation messaging during Box for Salesforce setup.

.. and other minor improvements and enhancements to the integration related to error messaging and usability.

For additional information, please refer to Box Support and Box for Salesforce documentation.

The misconception

If you've ever connected ChatGPT, Claude, or a custom MCP client to your Box account, you've probably thought: "As long as I block downloads, I'm safe. The agent can't take the file out."

The reality is harder. AI agents don't need to download a file to exfiltrate it. If an agent can preview a document, it can read it, comprehend it, and reproduce what's inside. From a data loss prevention perspective, an LLM reading content is functionally the same as downloading it. The moment your existing Integration Restriction stops at the download boundary, you have a gap.

Why this matters now

Enterprise workflows are moving from thousands of human users to a world where every user brings a handful of AI agents along with them. IDC projects that by 2028, roughly 70% of AI-driven enterprises will operate multi-tool, multi-agent architectures. That is a scale shift that human-era governance assumptions were not built for.

The structural problem is this: the controls that kept content safe from human misuse were designed around a click, a download, and an audit log. Agents don't click. They call APIs, they read text representations, and they answer questions on behalf of a user in seconds. Governing what an agent can read is now as important as governing what it can download, and it has to be enforced at the same layer, not bolted on later.

Architecture walkthrough

Architecture walkthrough

Classification-Based Access Policy extends the Integration Restriction security control in Box Shield with a new Read action, evaluated at request time against the classification on every file. Here is what happens end-to-end, stage by stage, with the security checkpoint named at each step.

A. Box user initiates access through a connected integration

A Box user initiated an action through a connected third-party integration. The integration sends the file requests to Box and identifies itself with an external serviceId. Box evaluates the integration’s serviceId to determine whether the requested action is allowed, blocked, or monitored.

B. Box Platform: SIS holds the rules

Shield's rule store (SIS) is the source of truth for every access policy an admin has configured. When an admin creates a Classification-Based Access Policy, the Download and Read toggles, the scope (block-all-except or allow-only), and the integration lists are persisted here. SIS also carries the enforcement mode for each action independently: enforce or monitor.

C. Box Platform: the enforcement engine decides

Every file access request from the connected integration lands in the enforcement engine. It pulls the file's current classification, looks up the restriction policy that applies to that classification in SIS, and evaluates the caller's serviceId against the Download and Read scopes independently. The policy evaluates the connected integration’s serviceId rather than the Box user alone. When a user accesses content directly, there is no restricted integration serviceId to block. When the user acts through a connected integration, the integration can be stopped by the Read restriction. Download and Read are two separately toggle actions on the same control, so an admin can block all integrations from previewing Confidential content while allowing a specific list to download, or any other combination.

D. Observability: every decision is logged

Every allow, block, and would-have-blocked (monitor mode) decision is emitted to the Event Stream API with the application ID, application name, action type (download or read), enforcement mode, file ID, user, and timestamp. Admins get full visibility into restriction activity across both action categories, without needing to instrument anything themselves.

E. Governed output: what the caller actually sees

When Read restriction fires, the integration receives a response with the classified content removed. It does not get a partial answer. It does not get the text representation.The caller can see the file exists (metadata is still returned today), but cannot ingest the content.

Architecture diagram

An enterprise scenario

Imagine a mid-market financial services firm. The compliance team classifies every deal document as Confidential the moment it lands in the M&A folder. A deal analyst uses ChatGPT as a connected integration to access their Box account and summarize public research faster.

Without Classification-Based Access Policy, the analyst opens the ChatGPT plug-in and asks it to summarize a document from the M&A folder. ChatGPT calls the preview endpoint, receives the text representation, and returns a clean two-paragraph summary of the target's non-public financials. The file was never downloaded. The content is now in the model's context.

With Classification-Based Access Policy enforced, the same request lands in the enforcement engine, is matched against the Confidential access policy, and is blocked at the read layer. ChatGPT replies that it cannot access the file. It proposes two workarounds: open in Box preview, or fetch the raw file. Both hit the same enforcement engine and both fail. The analyst still has their agent for public research. The M&A content stays inside the compliance boundary.

Questions your enterprise should ask

Questions your enterprise should ask

1. Can you tell, right now, which third-party integrations have ever previewed a Confidential file in your Box tenant?
2. If a user connects a new AI agent tomorrow, is your default posture "allow" or "deny" for reading classified content?
3. Do your integration policies distinguish between Download and Read as separate actions, or do you treat them as one?
4. Can you audit every read decision (blocked, allowed, or monitored) with the integration identity attached?

For most organizations, the honest answer to at least two of these is "not yet." Classification-Based Access Policy answers each of them directly. Read is a first-class action. Read restriction defaults to off, so no existing policy changes silently. Every decision is logged with the integration identity. And the same classifications you already trust drive the enforcement.

What you can build once this is governed

What you can build once this is governed

Because Read is now governed at the classification layer, you can safely:

• Roll out the Box MCP server to more teams, with confidence that Confidential and Internal content stays inside your rules.
• Approve new AI agents as they emerge without rewriting policy per app. Add the integration to the right Shield list, and the classification-based rules apply automatically.
• Adopt in monitor mode first. Run Read restriction in monitor for a classification to see who would have been blocked, then flip to enforce when you are ready.
• Extend the same identity-plus-classification pattern into your Event Stream consumers, so your SIEM sees agent activity the same way it sees user activity.

Bottom line

Bottom line

Classification-Based Access Policy is not a wrapper around the existing download control. It is a rebuild of the Integration Restriction enforcement path that treats Read as a first-class, independently governed action, tied to the classifications your business already trusts. One control, two independent toggles, every decision logged, and no path around it via preview, search, or text representation. That is what governing AI agents at enterprise scale looks like.

We're excited to announce Box for ServiceNow, a native, ServiceNow-certified integration that puts your Box content directly within ServiceNow records where work happens. It's available today on the ServiceNow Store.

Process lives in ServiceNow. Content lives in Box. Connecting the two isn't new; teams have long been able to automate how Box files move through ServiceNow workflows behind the scenes, without manual copying. But even with that automation in place, what's been missing is the record itself; the person working a ticket, case, or contract had no native way to open, preview, or share the relevant Box content without leaving ServiceNow and switching to Box.

Box already holds the files that work depends on. Box Intelligent Content Management now connects directly to ServiceNow. With Box for ServiceNow, content appears right on the ServiceNow record—no switching systems, and no content ever leaving Box's governance.

Your content, on the record

Box for ServiceNow brings content to people instead of sending them to Box and back. A Box panel sits on the ServiceNow record, so teams can navigate folders, preview files, upload, and share content in the same place they manage the work. Every record gets its own Box folder, created automatically, across ITSM, HR service delivery, CSM, CRM, or any custom module, so the right content is always attached to the right record.

Governance that stays with your content

Reaching your content shouldn't mean loosening your grip on it. Because your files stay in Box rather than being copied into ServiceNow, they keep every policy you already have on them: version control, classification, legal hold, and retention. Box remains the source of truth for your content, and ServiceNow remains the system of record for your process. Your governance follows the file — FINRA, HIPAA, SOC 2, ISO 27001, and FedRAMP included.

You can assign Box roles to ServiceNow users, so people see only what they're authorized to access, and no one can reach more content than they're entitled to in Box.

No custom code to maintain

Box for ServiceNow is a certified scoped application on the ServiceNow Store, so there's no custom-coded connector or third-party middleware for your teams to build and maintain. As Box and ServiceNow release updates, the integration keeps working, with no internal API project to maintain and no custom glue code to rewrite.

Built for document-heavy teams

The teams that feel the two-systems split most are the ones with the most content tied to every process:

• IT service management can keep the logs, root-cause analyses, and evidence a ticket depends on right on the incident.
• HR service delivery can keep onboarding paperwork and policy documents on the employee case.
• Legal and procurement can review contracts on the service request, with every file staying under Box's classification and retention rules.
• Customer support can put the documentation an agent needs on the case record.

Available today

Box for ServiceNow is available now on the ServiceNow Store. Your ServiceNow admin installs it in minutes — there is no middleware and no integration project, just a short setup to connect your Box account and configure access. It’s available at no additional charge for joint Box and ServiceNow customers on Box Business plans and above.

The integration is built to extend, and we'll keep bringing more of Box into the ServiceNow workflows your teams use every day.

[Get it on the ServiceNow Store] · [Read the setup and configuration guide]

ServiceNow, the ServiceNow logo, and other ServiceNow marks are trademarks and/or registered trademarks of ServiceNow, Inc. in the United States and/or other countries.

Legal work depends on documents, but finding, understanding, and acting on the right information can be time-consuming. Contracts, case files, matter histories, correspondence, policies, and research distributed across folders and systems make it difficult to bring the full context of a matter together.

Today OpenAI is introducing new offerings for law firms and the legal technology ecosystem, and Box is proud to contribute as a plugin partner and provider of legal skills. Together, Box and OpenAI are combining OpenAI’s intelligence with Box’s governed content foundation and legal expertise to help legal professionals turn enterprise content into practical, secure workflows.

Putting legal expertise to work with Box skills

The Box plugin includes skills designed around some of the most time-intensive workflows in legal work. Each combines ChatGPT’s reasoning capabilities with Box’s governed content, permissions, and collaboration controls.

• Virtual data room creation and management: Legal teams can use ChatGPT to create a secure data-room structure in Box, populate it with relevant files, and assign roles and permissions. The workflow can also support collaboration with outside counsel through controlled access and upload-only links — helping reduce setup time and permissioning errors.
• Client intake and onboarding: ChatGPT can review incoming client materials in a Box intake folder, check submissions for completeness, and perform preliminary risk assessments. Once the intake is approved, it can trigger the generation of an engagement letter, store it in Box, and share it through a secure workspace.
• Contract review and compliance: ChatGPT can scan executed agreements, identify deviations from firm templates, and assess potential risks as High, Medium, or Low. Results can be written back to Box metadata, creating a searchable view of contract risk without manual spreadsheet tracking.

Together, these skills help legal teams move their AI usage beyond document Q&A to supporting multi-step workflows that continue in Box, where content remains organized, governed, and managed.

Expanding the Box plugin with new legal skills

Box is also exploring additional skills for corporate legal departments and law firms.

• Outside counsel guideline management: Legal teams can use Box as their governed source for client-specific guidelines covering billing, staffing, security, data handling, AI usage, matter management, and reporting. ChatGPT can locate the applicable guideline, summarize key obligations, compare requirements, and surface relevant guidance in the context of a matter.

These skills turn Box’s legal subject-matter expertise into reusable workflows, helping legal professionals get more value from the content that’s already at the center of their work.

The governed content foundation for legal AI

Legal teams need AI that can work with sensitive enterprise content without weakening the controls around that content. Box provides that secure foundation.

Organizations can continue organizing, managing, and governing content in Box while using AI to analyze documents, surface relevant information, and support legal workflows. Existing Box permissions remain central, so users can work only with content they’re authorized to access.

That’s the value of Box as the file system for AI: enterprise content can be connected to whatever applications, models, and agents that legal teams choose, while Box remains the place where that content is organized, governed, and managed.

What this means for customers

The value of these experiences isn’t simply answering questions about documents; it’s helping lawyers turn Box content into meaningful legal work.

A transaction team can use Box to organize relevant files, manage permissions, and collaborate securely with outside counsel. AI can then analyze that content to extract key information and prepare a structured summary for review. A legal department can compare agreements stored in Box against internal requirements, surface material deviations, and write risk assessments back to Box metadata for ongoing visibility.

Where OpenAI provides the intelligence to reason across complex legal work, Box provides the governed content foundation where that work begins, continues, and is managed.

Building the future of legal work

Together, Box and OpenAI are helping legal professionals do more with the information that’s already at the center of their work. You can get started connecting ChatGPT to Box here.

The new AI-first era of business is here. Financial services firms have a greater need than ever to automate processes, extract insights, enhance collaboration, and protect valuable content. Today, that means using governed AI to search, analyze, extract, automate, and protect enterprise content across banking, insurance, wealth management, and M&A workflows.

With the power of Intelligent Content Management, teams can leverage their data (90% of which is unstructured!) to drive innovation and efficiency. For financial services firms, the opportunity is to turn unstructured enterprise content into trusted, actionable intelligence without sacrificing security or compliance.

Box, the leading Intelligent Content Management provider, delivers a secure, centralized platform with AI-powered features, enabling businesses to optimize workflow automation and ensure compliance. With innovations like Box AI, metadata extraction, and a new set of workflow automation capabilities, Box empowers financial services organizations to realize the full value of their content while addressing security concerns and enhancing operational effectiveness. This gives regulated teams a more secure foundation for applying AI to enterprise content — keeping documents, metadata, workflows, and permissions connected.

At BoxWorks 2025, Box announced new agentic AI, automation, and security capabilities designed to help financial services firms unlock more value from unstructured content while maintaining enterprise-grade governance.

BoxWorks 2025 highlights include:

• Box Extract uses AI agents to extract accurate, structured data from documents, presentations, images, scans, spreadsheets, and other content types.
• Box Automate orchestrates agentic workflows across AI agents, employees, and systems to reduce manual work in processes like onboarding, approvals, reviews, and invoice processing.
• Enhanced Box Apps brings AI-powered dashboards, metadata views, agent- assisted analysis, and natural-language exploration into no-code business applications.
• Box AI Agents and Box AI Studio help teams create purpose-built agents for tasks such as contract review, research, policy analysis, and client support.
• Box Shield Pro adds AI-powered content protection capabilities, including classification agents and agentic security insights for sensitive enterprise content.
• Enterprise Advanced brings together Box's most advanced content, AI, automation, governance, and security capabilities for organizations scaling AI across business workflows.

These updates expand the role of Intelligent Content Management for financial services: content is no longer just stored and shared, but understood, structured, automated, governed, and activated by AI across critical business workflows.

How do financial services firms use AI with enterprise content?

Financial services firms use AI with enterprise content to analyze documents, extract key metadata, summarize complex records, search across governed repositories, automate reviews, and trigger downstream workflows. In practice, this can include using AI agents to review loan applications, identify risky contract clauses, summarize client onboarding files, extract policy or claims data, generate personalized investment documents, and route approvals based on business rules.

The most effective approach keeps AI grounded in trusted enterprise content, applies role-based permissions, preserves auditability, and connects extracted data back to the original documents — so teams can work faster while supporting compliance.

This is especially important in regulated environments where sensitive content, customer data, transaction records, and long-term retention requirements must remain protected throughout the AI workflow.

We also announced Enterprise Advanced, which combines the full power of our Intelligent Content Management platform into a single offering. Let’s explore how our new capabilities enable financial services organizations to drive value with unstructured content.

Banking

In banking, our solutions streamline various processes by automating the extraction of metadata from submitted documents. This automated process replaces manual data entry and improves efficiency and accuracy in tasks such as Know Your Customer (KYC) verification.

By collecting key details like applicant names, submission dates, and loan types, our metadata extraction solution enhances searchability and reduces time spent searching for information. It also enables workflow automation by routing applications to the appropriate department based on criteria such as loan amount or applicant category. This speeds up processing times and promotes better collaboration within teams.

Integrating metadata into processing systems ensures compliance with regulatory requirements by accurately categorizing and securely managing sensitive information. Real-time access to crucial data facilitates efficient document generation for approvals and communications, resulting in faster response times and enhanced customer service. Leveraging form submission data, Box Relay workflow variables, and manual data entry can streamline processes such as generating preliminary loan agreements, while an AI-powered assistant can analyze loan applications for potential issues.

With additional features like Box Apps for tracking loan applications and organizing documents, banks can streamline their document management processes. A key advantage of Box Apps is its ability to leverage metadata collected from extract agents, enabling users to create custom views and dashboards. This allows employees to efficiently sort, filter, and access documents based on relevant metadata. Additionally, future capabilities like Box Archive will further enhance document management, ensuring long-term, compliant content preservation.

Insurance

For insurance firms, Box Forms enhances customer interaction and satisfaction around claim submissions by simplifying how policyholders detail incidents and attach supporting documents. Metadata extraction from customer-submitted forms and images enables insurers to seamlessly organize, retrieve, and process claims more efficiently—powering automation of critical workflows and streamlining the customer experience.

Based on the extracted data, Box Doc Gen facilitates the creation of acknowledgment letters confirming receipt of claims. Claims adjusters use a custom, metadata-powered dashboard in Box Apps to easily filter and sort submissions. After settlement, claims may be stored in Box Archive for long-term preservation. Additionally, teams can automate the creation of policy documents and renewal notices within their client databases or preferred system of record by leveraging the Box Doc Gen API, ensuring timely and accurate documentation.

Wealth management

In wealth management, Box Forms enables new clients to complete onboarding questionnaires, providing vital information about financial goals and risk tolerance. Metadata extraction processes customer financial statements, which can be stored in a firm’s CRM system.

Box Doc Gen then generates personalized investment proposals and account creation documents based on information submitted via a Box Form. Using Box AI Studio, financial analysts can build a custom AI agent that helps them analyze client profiles and empowers them to make informed decisions on investment strategies. All client-related documents are stored in Box Archive, which ensures client data is compliant, preserved — and protected with controlled access. Content can easily be restored if needed, with direct collaborations and original ownership restored as well.

The Box Doc Gen API seamlessly integrates data from portfolio performance and market analysts to create highly personalized investment reports. By leveraging this comprehensive information, gathered from various sources such as custom applications or Box Forms submissions, we enhance client engagement and empower informed decision-making.

Mergers and acquisitions

During M&A transactions, potential buyers submit information requests through Box Forms, detailing specific data needed from target companies. Metadata extraction pulls key details from company documents, identifying essential information such as financial figures and legal terms. While this organized data supports analysis and decision-making, the data remains within Box’s ecosystem rather than integrating directly into external project management systems.

Box Doc Gen then creates comprehensive due diligence checklists and document request lists using data collected from Forms submissions. An AI model evaluates responses against historical transactions to identify potential red flags. M&A teams create a tailored app to manage the due diligence process, ensuring alignment throughout. After completion, all related documents may be stored in Box Archive, ensuring the final version of content is preserved and unable to be altered, helping meet compliance requirements.

Our integrated solutions simplify and automate complex processes in banking, insurance, wealth management, and capital markets — so teams enhance efficiency and improve client experiences.

And that’s just the beginning of what Intelligent Content Management can do for financial services. Explore the latest BoxWorks updates to learn how content, AI agents, automation, and governance can work together to improve client experiences and operational efficiency.

Intelligent Content Management for financial services FAQs

What is Intelligent Content Management for financial services?

Intelligent Content Management for financial services is a secure approach to managing, protecting, automating, and applying AI to business-critical content — loan files, claims forms, client onboarding records, investment reports, and M&A documents. It helps firms connect content, metadata, workflows, and governance in one platform.

How can AI improve document workflows in banking?

AI can improve banking document workflows by extracting data from submitted files, identifying missing information, routing applications to the right teams, summarizing lengthy documents, and supporting faster review of KYC, lending, compliance, and customer service processes.

How do insurance companies use AI for enterprise content?

Insurance companies can use AI to organize claim submissions, extract details from forms and supporting images, generate acknowledgment letters, summarize claim files, and surface relevant policy information for adjusters — while keeping sensitive content governed.

Why is governance important when financial services firms use AI?

Governance is critical because financial services content often includes regulated, confidential, or personally identifiable information. AI should only access approved content, respect user permissions, maintain audit trails, and support retention and compliance requirements.

What role does metadata play in AI-powered financial services workflows?

Metadata gives AI and automation systems structured context about unstructured content. By capturing details such as client name, document type, account category, submission date, loan amount, policy number, or deal stage, firms can improve search, dashboards, routing, reporting, and compliance workflows.

How can financial services firms get started with AI and enterprise content?

A practical starting point is to identify high-volume content workflows, classify the sensitive data involved, define access and governance rules, pilot AI extraction or summarization on a narrow use case, and then expand into automated workflows once accuracy, security, and compliance requirements are validated.

And that’s just the beginning of what Intelligent Content Management can do for your business. Dive into the BoxWorks keynote to learn more.

What’s changed

• If Claude and/or ChatGPT are currently set to Available, they will move to Added by Default.
• Users will see Ask ChatGPT / Ask Claude in the Preview sidebar without first installing the integration.
• Clicking the action takes the user to Claude or ChatGPT with the Box file ID and a file-type-specific sample prompt.
• Users still need to authenticate with their own Claude or ChatGPT account before they can use the action.
• Customers already using these integrations will automatically receive the new file-type-specific prompts; no reinstall is required.

What’s has not changed

• If an admin has disabled or blocked Claude or ChatGPT, either individually or through the global integrations setting, that configuration will continue to be respected.
• Admins remain in full control and can disable either integration at any time under Admin Console > Integrations.

What if I'd prefer to keep things as they are?

You're always in full control. Admins can disable Claude or ChatGPT individually at any time from Admin Console > Integrations — that control isn't affected by this change. If you'd like to opt out before the rollout date, you can make that change any time before September 8, 2026.

When does this take effect? This change goes live on September 8, 2026.

How to Opt Out

• To disable a specific action: Admins can turn off Ask ChatGPT or Ask Claude individually by clicking Configure on the integration in Admin Console > Integrations and toggling each file-type action off.
• To opt out of Added by Default: Admins need to disable the integration completely, which removes user access rather than just the default status.

Access your Box content without leaving ChatGPT

A new embedded Box experience makes it easy to find, preview, and work with your enterprise content in ChatGPT.

AI is only as valuable as the context behind it. For businesses, much of that context lives in unstructured content—from terms buried in contracts to decisions captured in strategy decks. Making the right information readily available in tools like ChatGPT helps teams work faster and produce more relevant, actionable results.

Now, to enable secure and seamless access to enterprise content, OpenAI and Box are expanding their partnership to bring Box directly into the ChatGPT Library and more surfaces across ChatGPT. This gives users instant access to trusted business context where they are already working with AI, without needing to search for files or move content out of Box.

With Box content fully embedded across ChatGPT, users will be able to browse their Box folder hierarchy, select or @mention files for chat context, preview content alongside their conversations, and use the Box MCP server to take supported actions - all without leaving ChatGPT. Teams can spend less time searching for content—or transferring it out of where it is already secured and permissioned—and more time using trusted business information to generate relevant insights and move work forward.

Access your Box content without leaving ChatGPT

When working with enterprise content, the use of AI has often required users to manually find, download, and move files between systems, creating friction around incorporating AI into everyday work. The new Box integration with ChatGPT removes that friction by connecting users directly to the content they already have access to in Box and making files accessible throughout ChatGPT.

The Box MCP server already enables ChatGPT to securely find and work with enterprise content through natural-language prompts. The new embedded experience builds on that foundation by bringing a user’s Box folder hierarchy directly into the ChatGPT Library.

With the embedded Box experience, users can:

• Browse their Box folder hierarchy or search for files directly from the ChatGPT Library
• Select Box files to use as context for conversations and projects
• @mention a Box file directly in any new or existing conversation
• Preview source files alongside ChatGPT while asking questions or performing analysis
• Edit Box Notes within that file preview without leaving ChatGPT or switching sessions
• Use supported Box MCP tools to take action on content

Built on Box’s robust security foundation, the integration does not change what users can access. Existing Box permissions continue to determine which files and folders are available, while Box admins can control which MCP server tools and actions users can invoke. Together, these controls make content easier to discover and put to work without compromising the security and governance already in place.

A headless file system for AI

As the file system for AI, Box provides headless access to that context across models, agents, and applications while remaining the place where content is organized and managed. The new ChatGPT experience brings this vision to life: users can work with their Box content through the ChatGPT interface while Box remains the consistent content foundation beneath the experience.

For customers, this creates a more flexible way to put enterprise content to work. Teams can adopt the AI experiences that best support their workflows without copying content into separate repositories or rebuilding their content foundation for each new application.

Whether a clinical researcher is synthesizing study materials, a financial analyst is examining quarterly performance, or a legal professional is reviewing an agreement, users can bring the relevant Box content directly into ChatGPT and keep the source material visible as they work.

Your content stays organized and managed in Box. Its value extends wherever work happens.

Get started today

The Box integration with ChatGPT is generally available today for all Box customers, with OpenAI rolling out access to users of ChatGPT’s paid tiers over the next few days. From ChatGPT, users can browse their Box folder hierarchy, select or @mention a file to start a new conversation, add Box files to an existing conversation, and preview source content alongside their work.

The Box MCP Server is available via the Box plugin to all Box customers. To get started, go to the ChatGPT Plugin directory to install the Box plugin.

Users can now securely browse, access, and work with Box content directly in ChatGPT, bringing enterprise knowledge into AI-powered workflows

REDWOOD CITY, Calif.--(BUSINESS WIRE) -- Box, Inc. (NYSE:BOX), the leading Intelligent Content Management (ICM) platform, today announced it is bringing enterprise content and AI together to power more intelligent, actionable ways of working in ChatGPT. With the new integration, users can securely access and engage with their most important business content stored in Box directly from ChatGPT, making it easy to discover the right information, generate insights, and create new content. By putting files in the ChatGPT Library, Box and OpenAI are providing a frictionless way to leverage enterprise knowledge while maintaining Box’s enterprise-grade security, permissions, and governance.

“The next era of enterprise AI will be defined by how effectively AI can understand the content and context that power a business,” said Aaron Levie, co-founder and CEO of Box. “The opportunity is to move beyond simply giving AI access to content and create experiences where the right files and information are at the center of how people interact with AI. By bringing Box and ChatGPT together, we’re making enterprise knowledge a seamless and secure part of the AI experience, opening up entirely new ways for people to engage with their content and get work done.”

“The next era of enterprise AI will be defined by how effectively AI can understand the content and context that power a business.”
Aaron Levie, Box CEO

“By bringing the files and context people already rely on in Box directly into ChatGPT, we’re making it easier to move from finding information to understanding it, developing new ideas, and taking action - all in one flow," said Sondra Batbold, Product Lead, Personalization and Context at OpenAI. "This is an important step toward making ChatGPT a more connected place to get work done, while respecting the permissions and controls people already have in place.”

Empowering ChatGPT with enterprise knowledge from Box

Enterprises have vast amounts of valuable business knowledge stored across documents, presentations, contracts, research, and other unstructured content that power their business. But putting that knowledge to work with AI has often required users to manually find, download, and move files between systems, creating friction around incorporating AI into everyday work.

The new Box integration with ChatGPT removes that friction by connecting users directly to the content they already have access to in Box and making files accessible throughout ChatGPT. Content remains organized, managed, and governed in Box, while users can bring the information they need into their flow of work with ChatGPT. Box’s granular permissions ensure that users and AI can only access the information they are authorized to use, giving enterprises a secure foundation for incorporating AI into content-driven work without sacrificing existing security and governance controls.

“By bringing the files and context people already rely on in Box directly into ChatGPT, we’re making it easier to move from finding information to understanding it, developing new ideas, and taking action - all in one flow.”
Sondra Batbold, Product Lead, Personalization and Context, OpenAI

With Box and ChatGPT, users can now:

• Navigate their Box folder structure and browse, search, select, and mention content directly in ChatGPT to quickly find the right information without switching between applications;
• Engage with Box content by tagging specific files as context in chats and projects, previewing source files, and editing Box Notes directly in ChatGPT to seamlessly move from information to creation;
• Act on Box content using supported tools through Box’s MCP server to securely reference, extract, update, and create content and power more advanced, content-driven workflows.

The integration brings the context held in Box into a wide range of business workflows, including:

• Clinical researchers can accelerate research and analysis by @mentioning a clinical trial protocol stored in Box directly in a ChatGPT conversation, adding supporting study reports as context, and using those sources to synthesize findings and develop research briefs;
• Legal professionals can streamline contract review by selecting the specific agreement, legal playbook, and supporting documents from Box they want to use as context in a ChatGPT project, ensuring analysis is grounded in the right source material as they identify key provisions, assess risk, and suggest edits;
• Financial analysts can analyze quarterly results by bringing reports, forecasts, and performance updates from Box into ChatGPT and viewing source files side-by-side with their conversation as they investigate trends, validate findings, and develop financial reports;
• Insurance auditors can use Box’s MCP server to power more advanced review workflows that securely reference policies, claims files, and supporting documentation stored in Box to extract relevant information, identify inconsistencies, and update audit materials;
• Production managers can bring scripts, schedules, budgets, and other production materials from Box into ChatGPT to surface key updates and then edit a Box Note with the latest production brief directly in ChatGPT without switching applications.

Box and OpenAI working together

Today’s announcement builds on Box and OpenAI’s long standing partnership to bring enterprise content and AI together. Box customers can already leverage OpenAI’s latest models through Box AI to analyze, understand, and generate value from their content, while developers can build custom AI agents and applications that securely work with Box content using OpenAI’s developer tools. Together, these integrations give customers and developers greater flexibility to put enterprise content to work with AI.

Availability

The Box integration with ChatGPT is generally available today for all Box customers, with OpenAI rolling out access to users of ChatGPT’s paid tiers over the next few days.

For more information, visit the Box blog.

What's new

We're excited to introduce a new layer of security controls for the Box MCP Server. Box admins can now control which Box MCP Server tools each unpublished platform app is allowed to use. For every app—regardless of authentication method—admins can apply the organization-wide MCP policy or create a custom set of permitted tools specifically for that app. This gives organizations the same granular governance for their internal, custom-built apps that was previously available only for published integrations.

Every platform app listed in Admin Console > Platform > Platform Apps now includes its own Box MCP Server tab. From there, admins can define exactly which tools that specific app is allowed to use, independently of the organization-wide global controls.

To find it: go to Admin Console > Platform > Platform Apps, locate the app by name, click to open it, and select the Box MCP Server tab.

Admins can now:

• Choose between Global and Custom configuration for each unpublished platform app. Global inherits the organization's global MCP settings; Custom lets you define a set of tools scoped to that app alone.
• Enable or disable individual tools within a Custom configuration, grouped by category — such as Files and Folders, Search, or Collaboration.
• Grant a trusted internal app broader or different access than the global default without affecting any other app.
• Revert to the organization's global policy at any time by switching from Custom back to Global.

Why it matters

Because unpublished apps are typically internal or custom-built and haven’t undergone Box’s app-publishing review, admins need more granular control over what each app can access. These controls help enforce least-privilege access by ensuring each app can call only the tools it genuinely needs. They also give admins one consistent way to govern both published and unpublished MCP-connected apps across their organization.

Every department runs on content. Legal manages contracts, HR handles employee records, and marketing organizes creative assets. But while the work of each line of business needs to happen inside of documents, the process of managing the documents often happens all over the place.

Information gets copied into spreadsheets. Status updates live in Slack. Approvals move through email. Teams jump between file storage, workflow tools, databases, and custom apps just to move a document from one step to the next. When work is managed across all these external apps, workflows quickly become inefficient and hard to manage.

Box Apps was built to solve that problem. Instead of building custom software or stitching together multiple platforms, you can create purpose-built applications directly around the content your business already relies on. Then you can manage everything, from contracts and invoices to vendor and employee onboarding, all in one experience. And Box Agent now enables deep agent analysis across your content in Box Apps.

Key takeaways

• Box Agent in Apps is an autonomous reasoning agent that breaks complex goals into subtasks, searches content, and reasons across sources to deliver answers without opening files
• Enable deep, multi-document analysis entirely within Box, keeping content secure under Box’s enterprise-grade permissions model without requiring new tools or data movement
• By turning static documents into active, data-like assets, Box Agent in Apps transforms workflows into intelligent applications

Introducing Box Agent in Apps

Today, we’re announcing Box Agent in Apps, a deep agent capability built directly into the Box experience that handles the work that’s too complex, multi-layered, and document-intensive for any single search or automated rule.

Box Agent is neither a chatbot nor a search bar. It’s a multi-step, autonomous reasoning agent that breaks your goal into subtasks, searches across your content, reads documents, extracts what matters, reasons across sources, and delivers a high-quality, traceable answer without you opening a single file.

When you’re working in Box Apps and face a task that spans dozens of documents or requires cross-referencing, Box Agent can step in. AI does the investigation and provides an answer without a human spending hours sifting through content.

This is the difference between a content platform that stores your work and one that actively supports and scales it.

Legal scenario: New laws, existing contracts

Here’s a scenario that can happen within legal and compliance teams: A new law passes that changes data-sharing rules for employee benefits administration. You think this new regulation might impact your contract with your company's benefits provider.

Your benefits provider contract is buried somewhere in Box, alongside dozens of other vendor agreements, service arrangements, and compliance documents. You know the contract exists. You're just not sure exactly what it says about data sharing or whether it needs to be updated.

Answering that question manually means tracking down the right document, reading through dense legal language, and making a judgment call about whether the specific clauses in your agreement are affected by the new rule. And if your legal team needs to check other vendor contracts for similar exposure, that process multiplies quickly, turning a single compliance question into days of work, with real risk that something gets missed.

Here’s how Box Agent in Apps handles it:

1. Step 1 — You describe the task: Working inside Box Apps, you prompt Box Agent with the specifics of your situation and the text of the new law: "[Name law] just passed, and it affects employee benefits administration. [Text of law.] Find our benefits provider contract and tell me whether we should consider any changes to our contract in order to comply."
2. Step 2 — Box Agent understands the goal: The agent interprets the intent of your request, not just the keywords. It determines what kind of task this is (a targeted compliance review of a specific vendor agreement), what information is needed (contract clauses related to data sharing, data handling obligations, and employee benefits), and how to approach the workflow.
3. Step 3 — Box Agent searches and reads your content: It searches your Box App content for the benefits provider contract, retrieves the document, and reads through it, extracting the specific clauses that relate to data sharing, data handling obligations, and compliance requirements. If similar language appears in other vendor agreements, it surfaces those too, so nothing gets missed.
4. Step 4 — It reasons across the contract and the new law: Box Agent doesn't just summarize the contract. It applies the requirements of the new regulation as a lens against the specific language in your agreement. It identifies which clauses conflict with the new requirements, which provisions are silent on the relevant obligations, and which parts of the contract are already compliant.
5. Step 5 — You get a clear, actionable answer: The output is a structured summary: the specific clauses in your benefits provider contract that need to be updated, the language that's problematic, and the provisions that are already in good shape. You also get “the proof” — direct links back to the source document so your legal team can verify every finding.

What would have taken hours of manual contract review now takes minutes. And because it all happens inside Box, where the contract lives, there's no data movement outside your Box environment and no new tool to configure.

Why Box Agent in Apps changes the game

Box Apps has always been the place to view, manage, and organize your content. Box Agent makes it the place where complex work gets done, fast.

You stay in one place. Box Agent is embedded in the Box Apps experience. It’s neither a separate tool nor a different interface. The same environment where you view, manage, and collaborate on content is now the environment where you can run deep, multi-document analysis.

Your content is already there. You don’t need to upload files, configure a new system, or point an AI tool at your data. Box Agent works on the content that’s already in Box.

You get answers, not just results. Standard search gives you a list of files to go read yourself. Box Agent gives you a synthesized, traceable answer with the source documents linked so you can verify every finding.

You stay in control. As Box Agent works, you see a clear, plain-language view of what it’s doing — which files it’s accessing, what it’s looking for, and where it is in the process. Transparency is built in.

Your security model doesn’t change. Box Agent operates entirely within Box’s enterprise-grade infrastructure. It only accesses content you have permission to see. Governance policies, role-based access controls, and compliance settings all apply to your existing content automatically. No new tool to configure, no data leaving your environment, no new security surface to manage.

The next evolution of Intelligent Content Management

For years, enterprise content management has focused on storing files. Box Apps changed that by turning content into intelligent business applications. Box Automate went further by automating the workflows around that content. With Box Automate, your content isn’t just organized, it’s moving forward automatically. You can route tasks, extract data, and trigger next steps without manual intervention.

Box Agent is the next step. With Box Agent in Apps, your documents become data-like. Documents can then be incorporated into automated workflows. And your workflows become intelligent applications that can reason, investigate, and deliver answers at a depth and scale that wasn’t possible before.

The best business application isn’t another system to manage. It’s the one that’s already built around your content.

Get started

Box Agent in Apps is available now. Talk to your Box account team to learn more about enabling Box Agent for your organization, or join the AI Content Workflows user group to connect with the greater community.