Opsin
AI-native governance and security platform for enterprise AI agents across approved tools.
Website: https://www.opsinsecurity.com/
Cover Block
Publicly reported
| Field | Value |
|---|---|
| Name | Opsin |
| Tagline | AI-native governance and security platform for enterprise AI agents across approved tools [Opsin] |
| Headquarters | San Jose, United States [Crunchbase] |
| Founded | 2024 [Crunchbase] |
| Stage | Seed [Business Insider, October 2026] |
| Business Model | SaaS |
| Industry | Security |
| Technology | AI / Machine Learning |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (3+): James Pham, Oz Wasserman, Jeremy Mailen [Business Insider, October 2026] [Highperformr, 2026] |
| Funding Label | Seed, total disclosed about $21,000,000 [Business Insider, October 2026] |
Links
Publicly reported
- Website: https://www.opsinsecurity.com/
Summary and Signal
PUBLIC Opsin sells an AI security and governance layer for enterprise agents, and it merits investor attention now because the company has paired early customer logos with a fresh $21 million seed round at a moment when large enterprises are beginning to move from employee copilots to more autonomous agent workflows [Business Insider, October 2026] [Morningstar/Accesswire, October 2026]. Founded in 2024 in San Jose, the company appears to have grown out of a practical security problem, first around oversharing and access control in generative AI tools, then toward broader monitoring and governance as agent use cases expanded across approved enterprise environments [Crunchbase, Unknown] [YesPress, June 2026] [Opsin].
The product claim is straightforward: Opsin says it discovers agents, maps their access and intent, and helps security teams monitor behavior across tools such as Microsoft Copilot, Claude Enterprise, ChatGPT Enterprise, and Google Gemini [Opsin] [Morningstar/Accesswire, October 2026]. The differentiation, based on public materials, rests less on the model layer and more on visibility into what an agent can access, what it is meant to do, and whether its behavior diverges from that intended scope, which is a useful framing for regulated buyers if the company can prove low-friction deployment and actionable signal quality in production [Opsin] [Morningstar/Accesswire, October 2026].
The founding bench is relevant to the category. James Pham, listed publicly as co-founder and CEO, has prior experience at Abnormal Security and studied at MIT, while co-founder Jeremy Mailen is identified as CTO and previously co-founded Divebell before its acquisition by Exterro; public sourcing around Oz Wasserman confirms co-founder status, though his exact current title is inconsistently reported across sources [Risk Never Sleeps Podcast, June 23, 2026] [Crunchbase, Unknown] [Highperformr, 2026] [LinkedIn, retrieved 2026] [ZoomInfo].
On the commercial side, the public record is still early but directionally positive. Opsin says it serves nearly 20 enterprise customers, with named accounts including Culligan, Barry-Wehmiller, Wellstar Health System, and UiPath, and the company is selling on a SaaS model with hiring signals in both sales and engineering [Morningstar/Accesswire, October 2026] [Newswire, October 2026] [Business Insider, October 2026].
Financing is the clearest recent signal: Business Insider reported a $21 million seed round in October 2026 led by SYN Ventures and Race Capital, while startup database coverage also points to a $7 million corporate round in July 2025, though the latter is less fully documented in the available reporting [Business Insider, October 2026] [Indexed.vc, 2026]. Over the next 12 to 18 months, the public case will likely turn on whether Opsin can convert its current wedge into repeatable platform adoption across security, governance, and enforcement, while showing that its customer count can grow without diluting proof of efficacy in regulated environments [Morningstar/Accesswire, October 2026] [Opsin Customer Stories | Wellstar Health System, 2026].
Thinly sourced -- This section mixes independently reported funding and customer claims with several company-sourced product descriptions and partially corroborated team details.
Taxonomy Snapshot
| Axis | Value |
|---|---|
| Stage | Seed |
| Business Model | SaaS |
| Industry / Vertical | Security |
| Technology Type | AI / Machine Learning |
| Geography | North America |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (3+) |
| Funding | Seed, total disclosed about $21,000,000 |
Company Overview
PUBLIC
Opsin presents as a young enterprise security company built around a specific problem: giving security teams visibility into what AI agents can reach and do inside approved enterprise environments [Opsin]. Public records place the company in San Jose, founded in 2024, with Crunchbase classifying it as a seed-stage software business in the security category [Crunchbase].
The chronology that can be supported from the allowed public sources is still fairly short, which is typical at this age. Crunchbase attributes James Pham's start as co-founder and CEO to June 2024, anchoring the company's operating timeline, while the company website now positions Opsin as an "AI Agent Security Platform" focused on discovering agents, mapping intent, identity, data, and access across enterprise systems [Crunchbase] [Opsin]. Crunchbase also records a corporate round in July 2025 and a seed financing in October 2026, bringing disclosed funding to about $21 million in the current source set [Crunchbase].
One source, partially checked -- Confirmed by Crunchbase and the company website, with milestone detail limited to those two public sources.
The Product and the Stack
Public record plus analysis
Opsin is positioning around a problem that security teams are now encountering in plain sight: once enterprises approve tools such as Microsoft Copilot, Claude, ChatGPT Enterprise, and Google Gemini, they also need a way to see what those agents can reach, what they are intended to do, and whether behavior begins to drift from that mandate [Opsin] [Morningstar/Accesswire, October 2026]. Public materials describe the product as an AI agent security platform that discovers agents, maps intent, identity, data, and access, and evaluates tools and agents in context by comparing permissions, activity, and outcomes to surface risk signals [Opsin]. Press coverage repeats the same core workflow, discovery, monitoring, governance, and security for enterprise AI agents, with the company emphasizing regulated environments where data access and oversharing are harder to tolerate [Business Insider, October 2026] [Citybiz, October 2026].
The product surface described publicly is broader than a single monitoring dashboard, although most detail still comes from the company itself. Opsin's site and related company materials reference AI Agent Discovery, Agent Runtime Security, AI Agent Governance, MCP Governance, Agent Identity Security, AI Risk Assessment, and ongoing oversharing protection [Opsin]. A case study with Wellstar Health System says Opsin delivered a risk assessment within three days across petabytes of data and hundreds of Microsoft Teams and SharePoint sites, which is one of the few public implementation examples with any operational specificity [Opsin Customer Stories | Wellstar Health System, 2026]. There is also a free GenAI risk assessment offer, with results promised within 24 hours after onboarding, which suggests the company is using assessment-led entry points as part of the commercial motion rather than selling only a full platform deployment [Opsin].
The available public record does not support a detailed view of the underlying architecture, model layer, or deployment model, so any deeper technical claims would overreach the evidence. What can be said with confidence is narrower: Opsin is presenting itself as a control plane for enterprise AI usage, especially where agent identity, access rights, data exposure, and intended behavior need to be interpreted together rather than as separate security events [Opsin] [InfotechLead, October 2026]. Job postings for a senior AI engineer and a cybersecurity senior account executive indicate active hiring across both product engineering and go-to-market functions, but they do not justify firm conclusions about the stack beyond AI and cybersecurity orientation (inferred from job postings) [PERPLEXITY SONAR PRO BRIEF].
No independent source found -- Core product descriptions rely materially on company website and company-distributed materials, with partial repetition in Business Insider, Morningstar/Accesswire, Citybiz, and InfotechLead.
The Market They Are Entering
Publicly reported The market matters now because enterprises are moving from employee-facing copilots to more autonomous AI agents, and that shift turns a familiar data-access problem into a governance, identity, and runtime-control problem that security teams do not yet appear to have fully standardized around [Business Insider, October 2026] [Morningstar/Accesswire, October 2026]. Opsin’s public materials and financing coverage consistently frame the buyer need in those terms: enterprises are deploying agents across Microsoft Copilot, Claude, ChatGPT Enterprise, and Google Gemini, while security teams need visibility into what those systems can access, what they are intended to do, and when behavior changes in production [Opsin] [Business Insider, October 2026].
The available source set does not provide a named third-party TAM, SAM, or SOM for Opsin’s exact category, so any sizing work here has to stay at the level of analogous markets rather than direct market definition. Public reporting supports a narrower observation instead: the company is selling into enterprise security and governance budgets attached to approved AI tools and regulated data environments, with early customer references in healthcare, manufacturing, and other regulated settings [Morningstar/Accesswire, October 2026] [Yahoo Finance]. That suggests the practical near-term market is less "all AI" than the overlap between enterprise AI adoption, identity and data security controls, and governance requirements for agentic workflows [Opsin] [Business Insider, October 2026].
A useful way to read demand is through the problem sequence described in the sources. Opsin’s initial wedge was preventing sensitive-data oversharing and excessive access by generative AI tools, then expanding toward continuous discovery, behavioral monitoring, governance, risk assessment, and enforcement as enterprises adopt more autonomous agents [YesPress, June 2026] [Business Insider, October 2026]. That progression tracks a broader buying pattern in security: point concerns often open the door, but durable budgets tend to form when a new control plane starts to look operationally necessary across multiple tools and user groups.
The adjacent markets are clearer than the absolute size. Opsin appears to sit at the intersection of SaaS security posture management, identity governance, data loss prevention, insider-risk monitoring, and AI governance software, with some substitute pressure from incumbent cloud-security and collaboration-security vendors that can extend existing controls into Copilot, Gemini, or ChatGPT Enterprise environments [Opsin] [Business Insider, October 2026]. The distinction in Opsin’s pitch is that it evaluates AI tools and agents in context, comparing intent, permissions, activity, and outcomes, rather than treating these systems as only another SaaS application or another endpoint [Opsin].
| Market lens | Publicly supported read | Source basis |
|---|---|---|
| Core category | Enterprise security and governance for AI agents across approved enterprise AI tools | [Opsin] [Business Insider, October 2026] |
| Near-term buyer set | Enterprise security teams, especially in regulated or data-sensitive environments | [Morningstar/Accesswire, October 2026] [YesPress, July 2026] |
| Adjacent budgets | Identity, data security, governance, and runtime monitoring | [Opsin] [YesPress, June 2026] |
| Substitute spend | Existing security controls attached to SaaS, collaboration, and cloud environments | [Business Insider, October 2026] [Opsin] |
The table points to a category that is budget-adjacent rather than budget-isolated. For Opsin, that can help near-term adoption because the pain maps to existing security owners, but it also means the company will need to prove that agent-specific controls are distinct enough to avoid being absorbed into broader platform suites.
Regulation is part of the tailwind, though the public evidence is still mostly company-framed rather than independently quantified. Opsin’s materials explicitly connect the product to enterprise efforts to align with the EU AI Act, GDPR, and ISO/IEC 42001, and its cited customer mix leans toward sectors where governance language tends to move from policy discussion to procurement criteria faster than in less regulated environments [Opsin] [Morningstar/Accesswire, October 2026]. The strongest macro signal in the current record is not a formal market-size number, but the combination of a fresh $21 million financing, nearly 20 enterprise customers, and customer references spanning healthcare and manufacturing, which suggests investors and buyers are both treating AI-agent oversight as a real control gap rather than a speculative future need [Business Insider, October 2026] [Morningstar/Accesswire, October 2026].
One source, partially checked -- This section relies on two independent public reports for category demand and buyer context, but market sizing remains qualitative and product-market framing partly depends on company materials [Business Insider, October 2026] [Morningstar/Accesswire, October 2026] [Opsin].
The Competitive Field
MIXED Opsin appears to be positioning itself between established cloud and identity security controls on one side and newer AI-governance workflows on the other, with the pitch centered on governing what enterprise AI agents can access, do, and expose across approved tools such as Microsoft Copilot, Claude, ChatGPT Enterprise, and Google Gemini [Opsin] [Morningstar/Accesswire, October 2026].
The segment map is still forming, which is part of the opportunity and part of the risk. At the incumbent end, large security platforms already own adjacent control points in identity, SaaS posture, data security, and cloud monitoring, and those vendors can extend into AI oversight through policy, access, and activity data they already collect, an inference supported by Opsin's own framing around identity, data, tool, and behavioral context rather than a wholly separate control plane [Opsin]. At the challenger end, Opsin is trying to define a narrower category around AI agent discovery, runtime security, governance, and oversharing protection, beginning with approved enterprise AI surfaces rather than open consumer usage [Opsin] [Business Insider, October 2026]. The closest substitutes today may be internal security teams stitching together existing Microsoft, Google, and identity controls, especially for early deployments where the problem is visibility rather than full agent lifecycle governance, although the public source set here does not name those vendors directly enough to benchmark feature parity.
Where Opsin has an edge today is less about scale and more about fit for the immediate pain point. Public materials consistently describe an initial wedge in preventing sensitive-data oversharing and excessive access by generative AI tools, then expanding into continuous discovery, behavioral monitoring, governance, risk assessment, and enforcement for autonomous agents [YesPress, June 2026] [Opsin] [Business Insider, October 2026]. That progression matters because it suggests the company is not asking customers to buy an abstract AI-governance layer first, it is starting from a concrete security concern that regulated enterprises already recognize. The customer list also points in that direction: Wellstar Health System, Barry-Wehmiller, Culligan, UiPath, Encore Technologies, and Cascade have all been cited publicly, and Opsin said it served nearly 20 enterprise customers as of October 2026 [Morningstar/Accesswire, October 2026] [Yahoo Finance].
That edge is real but not automatically durable. Much of the current advantage appears to rest on product focus, founder-market familiarity, and speed, including leadership ties to Abnormal Security and a roughly 20-person team at the time of the October 2026 financing [Business Insider, October 2026] [Crunchbase]. Those are useful early advantages, but they are perishable if platform vendors fold similar controls into broader suites or if adjacent security products use their existing distribution into enterprise security teams to package AI-agent monitoring as an extension rather than a separate budget line. The public evidence does not yet show a proprietary network effect, a standards body position, or a public ecosystem lock-in that would make displacement especially difficult.
Opsin is most exposed wherever the control surface is owned by the platform itself. Microsoft, Google, Anthropic, and OpenAI each sit closer to the application layer for the tools Opsin says it supports, and existing enterprise security vendors may already control identity, endpoint, email, SaaS, or data-loss-prevention budgets that overlap with the same buyer conversation, according to Opsin's own product descriptions and deployment framing [Opsin] [Morningstar/Accesswire, October 2026]. In practice, that means Opsin has to prove that agent-specific context, intent mapping, and behavior monitoring produce risk signals that native admin consoles and general security telemetry do not. It also means the company may find it harder to push into categories where the buyer expects full-stack security consolidation rather than a specialist control.
The most plausible 18-month scenario is a bifurcation between specialist adoption in regulated enterprises and feature absorption by larger platforms. If enterprise customers keep deploying cross-vendor agents across Copilot, Claude, ChatGPT Enterprise, and Gemini, Opsin could be a winner because heterogeneous environments make a neutral control layer more valuable [Opsin] [Morningstar/Accesswire, October 2026]. If AI-agent oversight remains mostly native to a single platform per customer, the likely loser is the standalone category itself, because procurement can default to bundled controls from the platform that already owns productivity, identity, or model access. On today's public evidence, Opsin's path depends on whether multi-tool agent governance becomes a durable enterprise requirement before larger vendors make the category feel sufficiently covered inside existing suites.
Opportunity
PUBLIC
If Opsin executes, the prize is not a niche security tool but a control layer that large enterprises may increasingly treat as mandatory wherever AI agents are allowed to touch internal data, business systems, and regulated workflows [Business Insider, October 2026] [Morningstar/Accesswire, October 2026] [Opsin].
The clearest upside case is that Opsin becomes a default governance and runtime-security layer for enterprise AI agents across the major approved model environments. That outcome is reachable, not merely aspirational, because the public evidence already shows three things in place unusually early: a current enterprise product surface tied to Copilot, Claude, ChatGPT Enterprise, and Gemini [Morningstar/Accesswire, October 2026] [Opsin]; a customer set that includes healthcare, manufacturing, and software buyers such as Wellstar Health System, Barry-Wehmiller, Culligan, and UiPath [Newswire, October 2026] [Yahoo Finance]; and fresh capital, $21 million disclosed in October 2026, to add sales and engineering capacity at roughly the 20-customer and 20-person stage [Business Insider, October 2026] [Morningstar/Accesswire, October 2026]. In plain terms, the company is trying to sell into a real budget line before enterprise agent sprawl becomes harder to unwind.
The upside branches into a small number of identifiable growth paths, each dependent on a different catalyst.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Cross-platform control plane | Opsin becomes the independent layer enterprises use to discover, monitor, and govern AI agents across multiple approved model vendors and workflow tools | Enterprises standardize on more than one AI platform, raising the cost of point-by-point controls | Opsin already positions around supported environments including Microsoft Copilot, Claude, ChatGPT Enterprise, and Google Gemini, which suggests a cross-platform wedge rather than a single-vendor feature bet [Morningstar/Accesswire, October 2026] [Opsin] |
| Regulated-enterprise standard | The company wins a reputation as the preferred security layer for sectors where oversharing, access drift, and auditability matter most | A visible reference deployment or compliance-driven buying cycle in healthcare and other regulated industries | Public customer references already include Wellstar Health System, and the company explicitly frames its controls around governance and risk assessment tied to enterprise AI usage [Opsin Customer Stories | Wellstar Health System, 2026] [Morningstar/Accesswire, October 2026] [Opsin] |
| Land-and-expand from assessment to platform | A low-friction assessment product converts into broader runtime monitoring, governance, and enforcement across the customer estate | Fast time-to-value from an initial assessment, then expansion as more agents are deployed | Opsin publicly offers a free GenAI risk assessment and has described a Wellstar assessment delivered within days across large data estates, which supports a practical wedge into larger platform adoption [Free GenAI Risk Assessment] [Opsin Customer Stories | Wellstar Health System, 2026] |
The compounding logic is straightforward if the initial deployment works. A first assessment or discovery pass gives Opsin visibility into which agents exist, what they can access, and where behavior diverges from intended purpose; that in turn creates the preconditions to sell runtime monitoring, governance, risk scoring, and enforcement into the same account [Opsin] [Morningstar/Accesswire, October 2026]. Each new supported AI environment also makes the product more relevant to security teams that do not want separate controls for Copilot, Claude, ChatGPT Enterprise, and Gemini, and the named customer mix suggests the company is already testing this thesis in data-sensitive organizations rather than in low-stakes pilots [Morningstar/Accesswire, October 2026] [Yahoo Finance]. The early signs are still thin, but nearly 20 enterprise customers by October 2026 and active hiring across sales and engineering are at least consistent with a land-and-expand motion beginning to form [Morningstar/Accesswire, October 2026] [Business Insider, October 2026].
The size of the win depends on whether this category settles into a durable platform layer or remains a feature inside broader security suites. No credible public market-size dataset or directly cited public comparable is supplied in the record here, so precision would be false. Still, if the cross-platform control-plane scenario plays out and Opsin becomes a standard purchase for large regulated enterprises adopting autonomous agents, the outcome could reasonably support a multibillion-dollar enterprise value over time (scenario, not a forecast), because the company would sit on a recurring security and governance workflow attached to a new application layer that its customers appear to view as material enough to fund now, well before the market is mature [Business Insider, October 2026] [Morningstar/Accesswire, October 2026]. That is the central upside case: not that Opsin wins all of AI security, but that it owns one of the narrower, more persistent control points created by enterprise agent adoption.
No independent source found -- This section relies on a mix of independent coverage and company-sourced product claims, with the upside framing and valuation discussion explicitly presented as scenario analysis rather than observed results.
Sources
Publicly reported
[Opsin] Opsin | AI Agent Security Platform | https://www.opsinsecurity.com/
[Crunchbase] Opsin - Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/opsin
[Business Insider, October 2026] This startup raised $21 million to clean up the AI agent mess hitting companies | https://africa.businessinsider.com/news/this-startup-raised-dollar21-million-to-clean-up-the-ai-agent-mess-hitting-companies/37e80nm
[Highperformr, 2026] Jeremy Mailen | Highperformr | https://www.highperformr.ai/people/jeremy-mailen
[Morningstar/Accesswire, October 2026] Opsin Raises $21M to Keep AI Agents on Task and Out of Trouble | https://www.morningstar.com/news/accesswire/1236364msn/opsin-raises-21m-to-keep-ai-agents-on-task-and-out-of-trouble
[YesPress, June 2026] James Pham - Securing the Enterprise AI Endpoint | Opsin | https://yespress.io/james-pham
[Risk Never Sleeps Podcast, June 23, 2026] Episode #238. Can AI Keep a Secret?, with James Pham, CEO and Co-Founder of Opsin | https://pod.wave.co/podcast/risk-never-sleeps-podcast/episode-238-can-ai-keep-a-secret-with-james-pham-ceo-and-co-founder-of-opsin
[LinkedIn, retrieved 2026] James Pham - Publicis Health Media | LinkedIn | https://www.linkedin.com/in/james-pham/
[ZoomInfo] Oz Wasserman | ZoomInfo | https://www.zoominfo.com/p/Oz-Wasserman/-
[Newswire, October 2026] Opsin Raises $21M to Keep AI Agents on Task and Out of Trouble | https://www.newswire.com/news/opsin-raises-21m-to-keep-ai-agents-on-task-and-out-of-trouble-22640364
[Indexed.vc, 2026] Opsin Funding Rounds | Indexed.vc | https://indexed.vc/startup/opsin
[Opsin Customer Stories | Wellstar Health System, 2026] Wellstar Health System | Opsin Customer Stories | https://www.opsinsecurity.com/customer-stories/wellstar-health-system
[Citybiz, October 2026] Opsin Raises $21M Led by SYN Ventures and Race Capital to Expand AI Agent Security Platform | https://www.citybiz.co/article/919256/opsin-raises-21m-led-by-syn-ventures-and-race-capital-to-expand-ai-agent-security-platform/
[InfotechLead, October 2026] VC funding deals: Opsin, Sirona Medical, Shamrock AI | https://infotechlead.com/tech/vc-funding-deals-opsin-sirona-medical-shamrock-ai-98905
[Yahoo Finance, 2026] Opsin Raises $21M to Keep AI Agents on Task and Out of Trouble | https://finance.yahoo.com/news/opsin-raises-21m-keep-ai-130000000.html
[YesPress, July 2026] Opsin: The Startup That Wants Your AI to Know Less | https://yespress.io/opsin
[Free GenAI Risk Assessment] Free GenAI Risk Assessment | https://www.opsinsecurity.com/free-genai-risk-assessment
Articles about Opsin
- Opsin's $21 Million Seed Secures the AI Agents Inside Copilot and Claude — The San Jose startup, founded by veterans from Abnormal Security, is selling governance and runtime security to CISOs whose companies have already deployed AI agents.