Dust vs Moveworks (2026)

Dust offers more than 70 ready made MCP connectors that let agents use other systems' tools. MCP servers, native and remote, are on both plans, with up to 5 remote servers on Business and no limit on Enterprise. Alongside MCP, Dust maintains native connectors to more than 20 data sources, among them Slack, Notion, GitHub, Google Drive, Confluence, Zendesk and Salesforce, with up to 3 on Business and no limit on Enterprise. The native connectors cover the main systems and MCP covers the rest.

Agents act as well as read, and credits are charged for tool use that includes actions in connected apps. Zapier, Make, n8n and Power Automate are available on both plans. Each integration belongs to a Space, so what an agent can reach through it follows the customer's own access model, not a blanket service account. Dust's inbound APIs, for outside systems calling Dust, are separate. The 70 and 20 counts are Dust's own, with no published directory listing them.

Sourcedust.tt/home/product, /home/pricing feature table and credits FAQread 2026-09-12

Hundreds of named enterprise systems are connected under one layer, across seven classes, and the Assistant takes action in them, from resetting access to updating CRM records. Service desk and ITSM tools include ServiceNow, Jira, Freshservice, Ivanti Service Desk, BMC Helix Digital Workplace, ManageEngine and Cherwell. HR and finance cover Workday, Coupa, SAP Ariba, SAP Fieldglass and Oracle, and Salesforce covers CRM.

Identity and access providers include Ping Identity, Duo and BeyondTrust. Content and collaboration tools are Box, Confluence, Quip, Google Drive, LumApps, Igloo, Interact, Axero and Brightspot. Mail, calendar and chat cover Microsoft Outlook, Exchange Online, Gmail, Google Calendar, Google Chat and 8x8, and knowledge and data sources include Databricks, Upland RightAnswers, ComAround and BrainStorm QuickHelp, with DocuSign for signatures.

The Assistant is linked to thousands of action plugins and resolves requests end to end, resetting access and provisioning tools for IT, updating CRM data for sales, and automating expenses, purchase requests and approvals for finance. Those are authenticated writes into systems Moveworks does not own, and the Reasoning Engine chooses each tool call at run time rather than having it wired in advance.

Sourcemoveworks.com/us/en/platform/ai-assistant and /platform/integrations; moveworks.com/us/en/platform/integrationsread 2026-09-12

Multi agent orchestration with scheduled and event driven triggers is included on both plans, and agents run on those schedules and triggers without a person present. Pods are persistent shared workspaces where people and agents work together on a topic, project or initiative, with shared conversations, tasks, files and agents in one place. Frames are interactive dashboards and apps that agents produce. Pods and Frames run on a shared virtual computer and automations.

Work is built from reusable, governable skills, agents and memory that the customer combines into workflows that compound with use. Deep research tasks that need complex, multi step orchestration and tool use consume more credits. A run can involve the people in a Pod, agents working in parallel, the connected systems each step reads and writes, and the model chosen for each agent. Dust does not say what happens when a step in a multi agent run fails or whether it is retried.

Sourcedust.tt/home/pricing feature table and /home/productread 2026-09-12

The Reasoning Engine plans each request at run time, breaking it into steps and choosing the tools that carry them out across enterprise systems. It is an agentic AI system that orchestrates enterprise-ready LLMs and built-in tools for AI agents.

It identifies the employee's overall goal, works out the steps needed and breaks complex issues into smaller parts, then selects the tools for each, automating processes with many steps under strict security protocols and real time governance. Moveworks describes this as automation across the enterprise that plans, executes and adapts as it goes, and the Assistant handles complex requests in depth with extended thinking. Plans are built at run time, not coded in advance.

An employee asks, the Reasoning Engine plans, and agents built or installed through Agent Studio and the Marketplace carry out the work, while admins govern from a central console and plugins act in the systems of record. Developers can extend the sequence by building and deploying agents that plan, reason and execute securely across enterprise systems, and one example covers IT, HR and finance in a single conversational request. ServiceNow, which acquired Moveworks, adds orchestration across departments on its own platform.

Sourcemoveworks.com/us/en/platform/reasoning-engine, /platform and /platform/ai-assistantread 2026-09-12

Scheduled and event driven triggers come with multi agent orchestration on both Business and Enterprise, and automations help power Pods and Frames. Dust's own connectors sync in real time, so a change in Notion, Slack or GitHub reaches Dust without waiting for a poll, and an agent can be triggered by it.

Agents are reached in the Dust workspace, including Pods where people and agents work in parallel, through a Chrome extension that brings agents into the tools people use, and through Slack, Notion, GitHub and Drive. Zapier, Make, n8n and Power Automate are included on both plans and can start a Dust agent from anything they can see. Dust does not list which event types each connector supports or whether customers can combine conditions, and does not explain its automations further.

Sourcedust.tt/home/pricing feature table, /home/product and docs.dust.tt welcome pageread 2026-09-12

The Assistant reaches employees without being asked, through nudges and broadcasts, on web, chat and mobile among other surfaces, in over 100 languages. It anticipates what comes next and sends proactive nudges for recurring or time-sensitive tasks. Employee comms, a separate capability, lets internal communications teams broadcast timely, targeted messages to all or some users so they never miss a reminder to act. Each broadcast is a scheduled send to a chosen audience.

Employees can reach the AI Assistant on web, chat or company portals, as well as intranets, browsers and mobile. The Conversations API embeds the Assistant in applications Moveworks does not own, and a dedicated AI Assistant on web ships as its own release. The Reasoning Engine acts on requests from any of these surfaces, and nudges key on recurring and time sensitive conditions.

Sourcemoveworks.com/us/en/platform/ai-assistant, /platform and /platform/internal-communicationread 2026-09-12

Agents answer through retrieval augmented generation over a corpus Dust keeps current. Dust maintains its own connectors instead of relying on a generic integration provider, and real time sync keeps agents working from current information. Search, query and extraction across all company data are included on the plans. Semantic search covers unstructured content, and Table Query combines semantic search with SQL for structured data, so one run can answer a numeric question from a warehouse and a policy question from a wiki. Image analysis is also supported.

What an agent can retrieve is limited by the asking user's permissions, because data sits in Spaces with configurable access controls and each integration belongs to a Space. Named sources include Slack, Notion, Google Drive, Confluence and GitHub, with more than 20 native connectors in all and anything else over MCP. Dust does not describe chunking, embedding, citations back to a source passage or how fresh a retrieved answer is.

Sourcedocs.dust.tt welcome page with dust.tt/home/product and /home/pricingread 2026-09-12

Answers come from the customer's own connected systems, filtered to what each asker may see, and a dedicated authoring product keeps that corpus up to date. Enterprise Search gives instant, context-aware answers across business apps, tailored to role, region, language and access level without relying on exact keywords or rigid filters.

The AI Assistant connects to the systems and content sources the customer uses, and a grounded knowledge system ties answers to the most up-to-date and trustworthy documentation available, with fact verification among the AI safeguards. Knowledge Studio generates grounded knowledge articles to fill content gaps and improve self service for FAQs, so the content grows over time, with gaps found from what employees actually ask. Retrieval respects permissions, so two people asking the same question can get different results. This is a retrieval structure over the customer's own estate that lasts between runs and scales far past any context window.

Sourcemoveworks.com/us/en/platform/enterprise-search, /platform/ai-assistant, /platform/ai-knowledge-management and /platform/securityread 2026-09-12

Agents can keep a memory of each person they work with. Once the Memory tool is added in the builder, the agent decides during a conversation what is worth keeping, and a user can also ask it to remember something, such as a preference for morning meetings.

Memories are private to each user and kept separately for each agent, they last until the user deletes them, and the user can browse them in the agent's Memory tab and delete any single item.

Within a piece of work, Pods hold the shared conversations, tasks and files for a project, and long conversations are compacted so the agent keeps the thread.

Sourcedocs.dust.tt Agent Memory, Tools and Context Compaction pagesread 2026-10-21

The Assistant uses four kinds of memory, and none is a lasting store per user, since conversation context is cleared every 24 hours. Semantic memory is knowledge of the content, entities and terminology the organization uses, shared across the organization rather than kept per user.

Episodic memory is billed as recalling past conversations with the user, but it is defined as awareness of the current conversation, meaning what questions and answers have been exchanged and what decisions have already been made. Context is automatically cleared every 24 hours from the reasoner, so anything older than a day is gone.

Procedure memory is knowledge of the tasks that can be performed and the business processes or rules to follow, meaning the plugin catalog and the customer's configured rules. Working memory is awareness of which operations are in progress and at what stage, with a framework that tracks variables such as business object IDs, which is run state inside a single execution. The context window is assembled fresh on each turn and the engine reads everything every turn, so nothing is carried in from an earlier store.

Sourcehelp.moveworks.com/agent-studio/agentic-ai/memory-constructs, /assistant-behavior-framework and /agentic-reasoning-engine/how-the-reasoning-engine-worksread 2026-09-13

Agents stop and wait for a person before an action that needs approval. Every tool on an MCP server carries a stake level, set in its Tools and Stakes settings, and a tool that asks pauses the agent until the user approves or declines that action, with medium stake approvals remembered per agent. Dust's own connectors come with the gate built in for changes, so publishing or deleting in Contentful, writing to Grafana, changing an Airbyte sync or running a paid Bright Data scraper all wait for a yes.

The request reaches people where they are, through a sound alert when an agent is waiting, an email with allow and decline links for agents started by email, and an API call that approves or rejects an action. A credit circuit breaker also pauses an agent at 600 credits on one message and asks whether to continue. Spaces and Groups decide beforehand which data, tools and systems an agent can reach.

Sourcedocs.dust.tt changelog (tool stakes and approval entries, January to October 2026) and API reference, validate an actionread 2026-10-21

Guardrails surround what the agents do. A content moderation system filters harmful or inappropriate content, prompt protection guards against prompt injection attacks, and query risk assessment identifies and mitigates risks in user queries. Fact verification and identity validation add further checks, with authentication and authorization handled by deterministic systems and source-of-truth methods rather than LLMs. The Reasoning Engine operates within scoped action boundaries enforced at the infrastructure level.

Each of these limits what the agent may do, but there is no person in the loop. There is no approval queue, checkpoint or escalation on a risk threshold, and no human role is named, so no one signs off before the Assistant resets access, provisions a tool, updates CRM data or completes a purchase request, all actions it takes end to end. Enterprise permissions and scoped access govern who can reach what, and governance from ServiceNow belongs to the acquirer's platform.

Sourcemoveworks.com/us/en/platform/security and /resources/blog/why-autonomous-ai-agents-need-new-enterprise-security-modelread 2026-09-12

Business and Enterprise both carry SOC 2 Type II, GDPR compliance is stated throughout, and HIPAA can be enabled. A Trust Center sits at trust.dust.com, with a vulnerability disclosure page at dust.tt/home/vulnerability. Business includes SSO with Okta, Entra ID and JumpCloud for 5 or more seats, on demand. Enterprise adds SCIM provisioning, audit logs and advanced security controls, plus custom data retention and a custom MSA and DPA, so Business customers have SSO and Spaces but no provisioning or activity trail.

On both plans, data is organized into Spaces with configurable access controls, so admins decide which teams can reach which data sources. Integrations with Notion, Slack and Google Drive belong to Spaces, and Groups set permissions. Data is encrypted with AES-256 at rest and TLS 1.2+ in transit. Prompts, documents and outputs are never used to train models, though third party model providers may keep data for up to 30 days for safety and abuse monitoring before deleting it.

Sourcedust.tt/home/pricing feature table, /blog/what-is-data-sovereignty, /blog/secure-enterprise-software-ai-powered-teams and docs.dust.tt admin guideread 2026-09-12

Moveworks publishes ten certifications with their scopes and auditor marks. The ISO/IEC set covers 42001 for AI management, 27001, 27017, 27018 and 27701, SOC 2 Type 2 spans security, confidentiality, availability and privacy, CSA STAR Level 2 is held at Gold, and FedRAMP, GDPR and CCPA complete the list.

Controls cover access to how user data is processed and stored, encryption at rest and in transit, masking of sensitive PII, and identity validation that uses deterministic systems and source-of-truth methods for authentication and authorization, securing access to systems and data without LLMs. The design follows zero trust, never trust and always verify, with least privilege access and responses scoped to each person's permissions.

Assurance work includes application security reviews, vulnerability scanning and external penetration testing, red team exercises against enterprise and AI systems, and a HackerOne bug bounty open to all researchers at every severity level. The certification reports sit behind a Whistic profile reached through a demo request or an account executive. Security that ServiceNow's trust and controls add belongs to the acquirer.

Sourcemoveworks.com/us/en/platform/security and /platform/ai-assistantread 2026-09-12

On Enterprise, audit logs record what agents do as well as what people do. Each agent run, each tool an agent executes and each trigger that starts a run is its own event, and a tool event names the tool, the MCP server and the data sources it reached, and says whether the AI drove the action or a person started it.

The log is time ordered and tamper evident, searchable and exportable as CSV, and it streams to Datadog, Splunk, Amazon S3, Google Cloud Storage or any HTTPS endpoint. On every plan, analytics break credit use down by agent, member, model, tool and Pod, and a per call export lists each model call and tool call behind the bill.

Sourcedocs.dust.tt Audit Logs, Audit Logs events reference, Analytics and the consumption export APIread 2026-10-21

Every system access and data event is logged for full visibility and can be reviewed to support compliance and incident response. The Reasoning Engine works within scoped action boundaries enforced at the infrastructure level, with detailed audit trails for compliance, so the agent's actions leave a retained, reviewable record.

Reasoning can be inspected per interaction, not only in aggregate, because the AI Assistant shows its work and a reader can follow how a request became an action. Employee Experience Insights gives program owners tools to identify issues, track metrics and improve the assistant with governance and analytics, and a self service admin console centralizes configuration and monitoring. Analytics that ServiceNow adds belong to the acquirer.

Sourcemoveworks.com/us/en/platform/ai-assistant and /resources/blog/why-autonomous-ai-agents-need-new-enterprise-security-modelread 2026-09-12

Customers on Business and Enterprise choose US or EU data residency, so a self serve customer at €24 a seat gets the choice. Enterprise adds single tenant deployment. The platform's code is public at github.com/dust-tt/dust, linked from the site footer, and Dust mentions self hosting without documenting how to do it. On EU hosting, data is stored on servers in the European Union, and model inference is routed to EU regions where available, which means not for every model, with Mistral named as the European provider.

Dust separates sovereignty, residency and localization, and notes that a European data center run by a US based company can still be compelled to hand data to American authorities. Dust is based in Paris, so that CLOUD Act argument works in its favor, and it says so. Ardabelle, a named customer, chose Dust on EU hosting and GDPR grounds for private equity deal documents. The specific EU region is not named.

Sourcedust.tt/blog/what-is-data-sovereignty and /home/pricing feature tableread 2026-09-12

Hosting is on AWS across seven named regions, with a commitment that customer data never leaves its region and a dedicated GovCloud environment for federal clients. The regions are US Commercial, US GovCloud and Canada in North America, EU Commercial and the UK in Europe, and Australia and Japan in Asia Pacific. All customer data remains in its region and is never transferred elsewhere. Moveworks GovCloud is a dedicated, secure environment built for government agencies and federal clients, backed by FedRAMP authorization. There is no on premises deployment.

Sourcemoveworks.com/us/en/platform/security and /solutions/ai-for-federal-governmentread 2026-09-12

Ready agents and a template library come with every workspace. The default agents run as they are and each does its own job: @dust answers from all the company data a user can reach, @deep-dive spends ten to thirty minutes researching across the knowledge base, data warehouses and the web, and @analyst answers questions about how the workspace uses Dust.

Admins can switch each one on or off. Templates cover common use cases, and choosing one opens Sidekick with the template loaded, which drafts the instructions and suggests tools and data for the team to adjust before publishing. The workspace gallery then fills with the agents a company's own builders make.

Sourcedocs.dust.tt Templates and default agents pages (@dust, @deep-dive, @analyst)read 2026-10-21

An AI Agent Marketplace at marketplace.moveworks.com offers a curated library of ready-to-use, pre-built, production-proof AI agents that are customizable and free to install, with quick connector setup and MCP. Each agent does work once selected, installs on its own through a stated path with no added license, and can be removed without affecting the rest.

Next to the catalog, Agent Studio lets developers build and deploy AI agents that plan, reason and securely execute actions across enterprise systems, and Knowledge Studio, Quick GPT and Brief Me ship as separately named units. Specialized AI Assistants, which create expert agents, are marked coming soon.

Sourcemoveworks.com/us/en/platform/ai-assistant, /platform/ai-agent-builder and marketplace.moveworks.comread 2026-09-12

Every plan includes more than 20 frontier models from OpenAI, Anthropic, Google, Mistral and DeepSeek, across the GPT, Claude, Gemini, Mistral and DeepSeek families, open source models among them. The customer chooses the model for each agent. No model is locked behind a higher plan, though more capable models may use more credits per message. Teams can pick the right model for each workflow and switch as their needs change.

Mistral AI, a French founded company, is the European option, and for work that must stay under European jurisdiction Dust offers EU hosted infrastructure with model inference in EU regions. Dust presents model choice as a way to avoid vendor lock in and geopolitical risk by controlling where data is processed and which providers are used, and calls being model agnostic a core feature. Usage analytics by agent and by model show what each model costs per workflow. Dust's MCP connectors and its Conversation and Data Source APIs are separate from model routing.

Sourcedust.tt/home/pricing model list and FAQ, /home/product, and /blog/what-is-data-sovereigntyread 2026-09-12

Several language models are routed by the Reasoning Engine on the customer's behalf, but no provider is named and no choice is offered. The engine orchestrates enterprise-ready LLMs and built-in tools, so routing is plural, done by Moveworks and disclosed. Safeguards protect the LLMs against hallucination, data poisoning and prompt injection, and no customer data is used to train global generative models. Moveworks historically ran on its own reasoning stack.

There is no model picker, admin setting or bring your own key path, no model assigned per agent and no individual provider named. Orchestration is Moveworks' decision, made per task, and Moveworks presents that as value on day one with no model tuning, prompt chaining or advanced engineering required. Support for any model on the ServiceNow AI Platform belongs to the acquirer.

Sourcemoveworks.com/us/en/platform, /platform/ai-assistant and /platform/securityread 2026-09-12

The platform has its own API. Business includes a Conversation API for driving conversations with agents from outside, and Enterprise adds a Data Source API that writes into the knowledge layer. Platform documentation at docs.dust.tt, linked from the site footer, offers an llms.txt index of every page in Markdown, with endpoints in OpenAPI.

The platform's code is public at github.com/dust-tt/dust, and a developer platform at dust.tt/home/solutions/dust-platform covers Dust Apps, described as an extensible, API first platform for custom integration and LLM orchestration. Programmatic use is metered at $0.01 per credit on Business and custom on Enterprise. Zapier, Make, n8n and Power Automate are available on both plans, and a Chrome extension brings agents into other tools. The MCP connectors run the other way, letting Dust call other systems' tools. No endpoint reference or SDK page is cited, and the OpenAPI specification is mentioned but not shown.

Sourcedust.tt/home/pricing developer tools table, docs.dust.tt welcome page and site footerread 2026-09-12

Outside applications can call the Assistant through the Conversations API, which extends it into another application, portal or intranet. Developer resources sit on four hosts, with a Developer Hub at moveworks.com/us/en/developers, help documentation at help.moveworks.com that includes a Moveworks Setup admin reference, a community at community.moveworks.com and training at academy.moveworks.com.

Agent Studio is a build surface where developers build and deploy agents and plugins that act across enterprise systems, and the Assistant is linked to thousands of action plugins. MCP ships as part of quick connector setup on the AI Agent Marketplace, so an outside assistant can reach Moveworks. The Conversations API is described on a product page, and Moveworks names no public reference for it.

Sourcemoveworks.com/us/en/platform/ai-assistant, /platform/ai-agent-builder and the developer estate linked in the site navigation; moveworks.com/us/en/developersread 2026-09-12

A builder can try a change before it goes live. A preview area beside the agent builder runs the draft agent on real questions, and the docs advise testing changes there until the answers are consistent.

Teams that manage agents as code can keep agent and skill definitions in YAML in a GitHub repository with the official GitHub Action, review edits in pull requests and promote them from a development workspace through staging to production, and changes Sidekick proposes from a conversation arrive as cards to accept or decline. Thumbs up and down on each answer are collected and readable through the API.

None of this scores a change, since there are no test cases, pass rates, evaluators or comparisons between two versions of an agent, so judging a change still rests on someone reading the answers.

Sourcedocs.dust.tt Troubleshooting your agent, the dust-github-action changelog entry and the feedback APIread 2026-10-21

Output is checked for accuracy at run time and admins get tools to tune the Assistant, but there is no way to test a change before employees see it. Fact verification improves the accuracy of generated information as a standing safeguard, alongside query risk assessment and a content moderation filter, all of which check what the agent produced.

Around it sits a tuning loop for the Assistant, as Employee Experience Insights lets program owners identify issues, track metrics and improve the assistant, and cutting-edge optimization tools let admins track metrics, resolve issues and make enhancements. There is no test environment, sandbox, holdout or simulation, and no versioning or scored result to compare.

Fact verification gives the customer no verdict to compare across versions, and nothing runs a new or edited agent against a known set before it reaches employees, though these agents reset access and provision tools.

Sourcemoveworks.com/us/en/platform/security, /platform and /platform/ai-assistantread 2026-09-12

Pods and Frames run on a shared virtual computer and automations, a shared environment where agents run code and produce artifacts, not one where an agent operates a graphical interface. Credits are charged for tools such as search, data retrieval, code execution and actions in connected apps, and interface control is not among them. The Chrome extension lets a person reach Dust agents while browsing, and no agent drives the browser.

Every path in and out of Dust is a maintained connector, an MCP server or an API, so a third party redesigning its interface does not affect it, and Dust maintains typed connectors instead of scraping. There is no hosted or local browser session, desktop or remote computer control, RPA, recorder or extension that acts on a page for the agent. Dust does not explain what its shared virtual computer can and cannot do.

Sourcedust.tt/home/product, /home/pricing credits FAQ and /home/chrome-extensionread 2026-09-12

The browser is a place the Assistant lives, not an interface it drives. Browsers are among the access surfaces alongside chat, intranets and service portals, and an AI Assistant on web ships as a destination employees open. On every path the employee opens the surface and types, and the Assistant answers and acts. Actions run through code, as the Assistant is linked to thousands of action plugins and the Reasoning Engine selects the right tools across hundreds of connected systems.

Resetting access, provisioning tools, updating CRM records and submitting purchase requests all run through those connectors and the Conversations API. Nothing clicks, fills in or advances a page, so a change in a target application's markup has nothing to break. There is no hosted or local browser session, desktop control, remote or local computer use, or RPA fallback.

Sourcemoveworks.com/us/en/platform/ai-assistant, /platform and /platform/ai-agent-builderread 2026-09-12