LONDON, May 18, 2026 — Paris-founded AI startup Dust has closed a $40 million Series B round led by Abstract and Sequoia, bringing its total funding past $60 million as it pursues a thesis that individual AI copilots are a dead end for enterprise productivity. The company, co-founded in 2023 by Gabriel Hubert and Stanislas Polu, now serves more than 3,000 organisations, recorded 41,000 monthly active users in April 2026, and has deployed over 300,000 agents across its platform. Strategic co-investors Snowflake and Datadog signal that the data infrastructure layer views Dust's "multiplayer AI" workspace — where humans and agents share context, notifications, and goals — as a viable new category distinct from both copilots and enterprise search. As Business20Channel.tv's enterprise AI coverage has tracked throughout 2026, the gap between individual assistant usage and measurable organisational productivity remains the central unsolved problem for CIOs. This analysis examines Dust's competitive positioning against Microsoft Copilot and Glean, the capital allocation logic behind Sequoia's bet, and the governance implications for sectors from finance to healthcare. For background on the broader enterprise AI investment cycle, see our Enterprise AI Funding Tracker 2026.

Executive Summary

• Dust raised $40 million in Series B funding led by Abstract and Sequoia, with participation from Snowflake and Datadog, taking total capital raised above $60 million.
• The platform connects to over 100 data sources and integrates with Slack, Salesforce, and Google Drive, deploying agents with built-in memory and feedback loops.
• Dust reported zero customer churn in 2025 and 70% weekly active usage among its 41,000 MAUs — metrics that distinguish it from most enterprise SaaS cohorts.
• Funds will target three priorities: self-learning agents, improved human-agent collaboration, and strengthened enterprise governance and audit systems.
• The company employs approximately 98 people across offices in Paris and San Francisco.

Key Developments

Funding and Investor Composition

The $40 million Series B represents a significant step-up for Dust, which was founded only in 2023. The round was co-led by Abstract and Sequoia, two firms with contrasting portfolio strategies whose joint leadership of this deal underscores the breadth of conviction behind the "multiplayer AI" thesis. Snowflake and Datadog's participation is particularly instructive: both are publicly traded data infrastructure companies (Snowflake, NYSE: SNOW; Datadog, NASDAQ: DDOG) that rarely take venture-stage positions unless a startup's product sits directly in their customers' workflow. With total funding now exceeding $60 million, Dust has accumulated meaningful capital relative to its 98-person headcount, giving it roughly $612,000 in cumulative funding per employee — a ratio that suggests the company is investing heavily in product engineering rather than sales headcount at this stage.

Product Architecture and the "Multiplayer" Thesis

Gabriel Hubert, Dust's CEO, described the core product philosophy in terms that directly critique the incumbent approach: "What will transform the way we work isn't the next best model or assistant. It's going to be a completely new type of system that gives humans and agents shared, governed access to the same information and capabilities so that they become true collaborators, working with the same context, notifications, artefacts, and goals to compound organisational impact. This is what we call multiplayer AI, and this is what we're building at Dust." The platform connects to more than 100 data sources and integrates with tools including Slack, Salesforce, and Google Drive. Agents deployed on Dust improve over time via built-in memory and feedback loops grounded in real team workflows. Governance features include detailed permissions, cost and usage tracking, full audit trails, and agent analytics — capabilities that enterprise procurement teams increasingly require before signing off on AI platform spend.

Operational Metrics

Dust's operational numbers are notable. The company had 41,000 monthly active users in April 2026 across more than 3,000 organisations. It has deployed over 300,000 agents on its platform — roughly 7.3 agents per active user on average, suggesting deep multi-agent deployment within customer organisations rather than surface-level experimentation. Dust reported zero churn throughout 2025, a claim that, if sustained, would place it among the most retentive enterprise SaaS products in any category. Konstantine Buhler, partner at Sequoia, quantified the engagement further: "Most enterprise AI today is single-player: one person, one prompt, no compounding. Dust is building a multiplayer system in which agents and humans share context and work together across the entire company. Zero churn and 70% weekly active usage tell you this isn't experimental anymore. This is how enterprises will actually operate." That 70% weekly active usage figure is a striking data point; for comparison, Microsoft 365 Copilot has faced public scrutiny over adoption rates within large enterprise deployments since its launch.

Market Context & Competitive Landscape

The Copilot Ceiling

Dust's pitch rests on a specific diagnosis of the enterprise AI market that splits incumbents into two categories, both of which it considers insufficient. The first category — foundation model workspaces and copilots such as Microsoft Copilot and Google Gemini for Workspace — helps individual users draft emails, summarise documents, and answer questions. However, according to Dust, these tools do not share context across teams or compound learning at the organisational level. Each user's interactions remain siloed. Glean, the enterprise search platform that has raised over $300 million to date, represents the second category: it can surface information but cannot take action on behalf of users or teams. Dust argues that neither approach solves what it calls "the compounding problem" — where individual AI improvements fail to aggregate into measurable organisational intelligence. This framing is intellectually compelling but carries risk: Microsoft and Google have the distribution, the data, and the capital to iterate their copilot products toward shared-context features over time.

Honest Assessment of Limitations

Dust's 41,000 MAU base, while growing, is minuscule relative to Microsoft 365's more than 400 million paid seats or Google Workspace's estimated 3 billion users. The "zero churn in 2025" claim is impressive but must be contextualised: the company's customer base was small enough in 2025 that losing even one mid-market account would have been a statistically significant event. Dust's platform relies on third-party foundation models rather than proprietary large language models, which means its differentiation must come entirely from the orchestration, governance, and collaboration layers — areas where well-resourced incumbents can invest aggressively. The 98-person team, while efficient, will face scaling challenges as enterprise customers demand dedicated support, compliance certifications, and regional data residency.

Industry Implications

Financial Services and Regulated Industries

Dust's emphasis on governance — permissions, audit trails, cost tracking, and agent analytics — positions it well for regulated verticals where AI deployment without oversight is a non-starter. In financial services, the European Union's AI Act, which entered its enforcement phase in early 2026, requires organisations to maintain detailed records of AI system usage, decision-making processes, and human oversight mechanisms. Dust's full audit trail feature maps directly to these requirements. Banks and insurance companies deploying AI agents for operations, compliance, and customer support need platforms that can demonstrate regulatory traceability — a capability that neither Microsoft Copilot nor Glean currently foregrounds.

Healthcare, Legal, and Government

In healthcare, shared-context AI agents could coordinate between clinical operations, billing, and patient communications teams while maintaining GDPR and HIPAA-compliant access controls. Legal teams, which increasingly adopt AI for contract review and case research, benefit from the multiplayer model because matters involve cross-functional collaboration between lawyers, paralegals, and clients. Government agencies face unique challenges: multiple departments must share intelligence and coordinate responses without breaching classification boundaries. Dust's detailed permissions system is architecturally suited to these use cases, though the company would need to pursue specific certifications (FedRAMP in the United States, for instance) to compete seriously in public sector procurement. The company has not yet disclosed plans for such certifications, which could limit its addressable market in the near term.

Business20Channel.tv Analysis

The Capital Allocation Logic

From an investor perspective, the $40 million Series B is a bet on category creation rather than category capture. Sequoia's Konstantine Buhler made this explicit when he stated: "Most enterprise AI today is single-player: one person, one prompt, no compounding." The implication is that Dust is not competing for share within the existing copilot or enterprise search markets — it is attempting to define a new segment that sits between them. This is a high-risk, high-reward strategy. If Dust's "multiplayer AI" thesis proves correct — that organisational productivity gains require shared agent-human context — then the company has first-mover advantage in a market that could be worth tens of billions of dollars by the end of the decade, based on Gartner's projections for enterprise AI platform spend. If the thesis is wrong, or if Microsoft and Google absorb the core insight into their existing products within 18 months, Dust's differentiation erodes rapidly.

What the Metrics Actually Tell Us

The 70% weekly active usage figure cited by Sequoia's Buhler deserves close scrutiny. In enterprise SaaS, weekly active usage rates above 60% typically indicate that a product has become embedded in daily workflows rather than sitting as an optional tool. For comparison, Notion reported approximately 65% weekly active usage among its enterprise customers in 2025, and Figma has historically reported rates in the 70–80% range among design teams. If Dust can sustain this engagement level as it scales beyond 41,000 MAUs, it will have a compelling retention story. The 300,000 agents deployed across 3,000 organisations suggest deep product penetration: teams are not merely experimenting with a single bot but building complex, multi-agent workflows. This density — roughly 100 agents per organisation — indicates that Dust's customers are treating the platform as infrastructure, not as a point solution.

Strategic Risks Worth Monitoring

Three risks stand out. First, Dust's reliance on third-party foundation models means that any significant pricing changes from OpenAI, Anthropic, or Google could compress its margins overnight. Second, the company's Paris-San Francisco structure, while giving it access to both European and American talent pools, creates operational complexity in an organisation of only 98 people. Third, the "zero churn" metric will face its real test as Dust moves upmarket into larger enterprise accounts where procurement cycles are longer, IT requirements are stricter, and switching costs are lower than in mid-market deployments. As our enterprise AI churn analysis has noted, the true test of retention comes when a platform's customer base exceeds 10,000 organisations and includes Fortune 500 accounts with complex vendor management processes.

Why This Matters for Industry Stakeholders

For CIOs evaluating enterprise AI platforms in 2026, Dust's Series B raises a practical question: should procurement teams wait for Microsoft and Google to add shared-context capabilities to their copilots, or should they invest now in a purpose-built multiplayer platform? The answer depends on risk tolerance and timeline. Organisations that need measurable cross-team AI productivity gains in the next 6–12 months may find that Dust's 100+ data source integrations and 300,000-agent deployment track record offer a faster path than waiting for incumbent roadmap promises. However, enterprises with heavy Microsoft 365 or Google Workspace commitments face integration overhead and potential vendor lock-in risk.

For AI product leaders at competing platforms, Dust's zero-churn claim and 70% weekly active usage rate represent a benchmark to match. If these metrics hold at scale, they will shift the conversation in enterprise AI procurement from "which model is best" to "which platform compounds organisational intelligence most effectively." Snowflake and Datadog's investment also signals that the data infrastructure layer sees multiplayer AI workspaces as a natural customer for their APIs — a channel partnership dynamic that could accelerate Dust's go-to-market in data-intensive industries. Venture investors tracking the enterprise AI space should note that Dust's $60 million total raise is modest compared to competitors like Glean, which has raised over $300 million, suggesting significant room for follow-on funding if growth metrics hold.

Forward Outlook

Dust's roadmap, as disclosed in connection with this funding, centres on three priorities: self-learning agents that improve without manual retraining, deeper human-agent collaboration with equal access to shared projects and information, and enhanced governance systems for enterprise-grade operations. The self-learning agent goal is the most technically ambitious and the most commercially important — if Dust can demonstrate agents that measurably improve over weeks and months through real team feedback, it will have a differentiation layer that pure copilots cannot easily replicate.

The next 12 months will be decisive. Gabriel Hubert's framing of "multiplayer AI" as a category must survive contact with enterprise procurement committees, IT security reviews, and the inevitable competitive response from Microsoft and Google. We expect Dust to announce its first major enterprise customer logos by Q3 2026 and to pursue SOC 2 Type II and potentially ISO 27001 certifications to unlock regulated industry sales. If the company can double its MAU count to 80,000 by early 2027 while maintaining its churn and engagement metrics, a Series C at a valuation exceeding $500 million would not be unreasonable. The open question remains whether "multiplayer AI" is a genuine new category or a feature set that incumbents will absorb — and only the next two years of enterprise adoption data will provide the answer. For continued coverage of enterprise AI platform funding and competitive dynamics, see Business20Channel.tv's AI intelligence hub.

Key Takeaways

• Dust's $40 million Series B, co-led by Abstract and Sequoia with strategic participation from Snowflake and Datadog, takes its total funding past $60 million and validates the "multiplayer AI" category thesis.
• With 41,000 MAUs, 300,000+ deployed agents, zero churn in 2025, and 70% weekly active usage, Dust's operational metrics outperform typical enterprise SaaS benchmarks at this stage.
• The company's competitive positioning directly challenges Microsoft Copilot's single-player model and Glean's search-only approach, but its 41,000 MAU base is a fraction of incumbents' install bases.
• Regulated industries — finance, healthcare, legal, government — represent high-value target verticals, but Dust must secure compliance certifications to compete for these contracts.
• Strategic risks include reliance on third-party foundation models, scaling challenges with a 98-person team, and the possibility that incumbents absorb multiplayer features into their existing platforms.

References & Bibliography

[1] TechFundingNews. (2026, May 18). Sequoia backs Paris AI startup Dust with $40M to fix the enterprise AI productivity gap. https://techfundingnews.com/dust-40m-series-b-collaborai-ai-enterprise-workspaces-sequoia/

[2] Dust. (2026). Official website. https://dust.tt/

[3] Sequoia Capital. (2026). Portfolio and investment thesis. https://www.sequoiacap.com/

[4] Snowflake Inc. (2026). Company overview and investor relations. https://www.snowflake.com/

[5] Datadog Inc. (2026). Company overview and investor relations. https://www.datadoghq.com/

[6] Microsoft. (2026). Microsoft 365 Copilot product documentation. https://www.microsoft.com/en-us/microsoft-365/copilot

[7] Google. (2026). Gemini for Google Workspace. https://workspace.google.com/products/gemini/

[8] Glean. (2026). Enterprise AI search platform. https://www.glean.com/

[9] European Commission. (2026). EU AI Act regulatory framework. https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai

[10] GDPR.eu. (2026). General Data Protection Regulation overview. https://gdpr.eu/

[11] Gartner. (2026). Newsroom — enterprise AI market projections. https://www.gartner.com/en/newsroom

[12] Bessemer Venture Partners. (2026). BVP Cloud Index. https://www.bvp.com/atlas

[13] OpenView Partners. (2025). SaaS benchmarking reports. https://www.openviewpartners.com/

[14] Salesforce. (2026). CRM platform and integrations. https://www.salesforce.com/

[15] Slack. (2026). Business communication platform. https://slack.com/

[16] Google Workspace. (2026). Google Drive and collaboration tools. https://workspace.google.com/

[17] Notion. (2026). Connected workspace for teams. https://www.notion.so/

[18] Figma. (2026). Collaborative design platform. https://www.figma.com/

[19] Microsoft. (2026). Corporate overview. https://www.microsoft.com/

[20] Google Cloud. (2026). Cloud platform and AI services. https://cloud.google.com/

[21] Business20Channel.tv. (2026). Enterprise AI intelligence hub. https://business20channel.tv/?category=AI