Escape
AI-powered offensive security platform for API security testing, attack surface management, and DAST.
Website: https://escape.tech/
Cover Block
Open sources
| Field | Value |
|---|---|
| Name | Escape |
| Tagline | AI-powered offensive security platform for API security testing, attack surface management, and DAST. |
| Headquarters | Paris, France [Crunchbase] |
| Founded | 2020 [Crunchbase] |
| Stage | Series A [Tech.eu, March 2026] |
| Business model | SaaS |
| Industry | Security |
| Technology | AI / Machine Learning |
| Geography | Western Europe |
| Growth profile | Venture Scale |
| Founding team | Co-Founders (2): Tristan Kalos, Antoine Carossio [TechCrunch, June 2023] |
| Funding label | $10M+ |
| Total disclosed funding | ~$22,900,000 [Startup Savant, October 2024] [TechCrunch, June 2023] [Tech.eu, March 2026] |
Links
Open sources
- Website: https://escape.tech
- LinkedIn: https://www.linkedin.com/company/escape-tech
- Y Combinator: https://www.ycombinator.com/companies/escape
What an Investor Needs First
PUBLIC Escape sells application security software that tests APIs and modern applications the way an attacker would, and it merits attention now because the company has progressed from an API-focused wedge to a broader AI-positioned offensive security platform while attracting a fresh $18 million Series A led by Balderton Capital in March 2026 [TechCrunch, June 2023] [Tech.eu, March 2026]. Founded in 2020 and headquartered in Paris, Escape appears to have started from a narrow but real technical gap: the founders focused on API business-logic vulnerabilities, an area that traditional scanners often handled poorly, and later expanded that premise into a wider platform story around pentesting, attack-surface management, and DAST [TechCrunch, June 2023] [YesPress] [Escape].
The product claim is straightforward even if the current scope is still largely company-described: Escape says its platform automates business-logic testing, agentic pentesting, code-to-cloud correlation, attack-surface management, and DAST, with the original product built to generate realistic requests, simulate attacks, surface flaws, and suggest remediations [Escape] [TechCrunch, June 2023]. That positioning matters because it frames Escape less as a point API scanner and more as an attempt to compress manual offensive security work into a continuous software workflow, although public evidence on efficacy, retention, and deployment depth remains limited [Escape job posting] [Tech.eu, March 2026].
The founding team is technically credible on paper. Tristan Kalos is identified as co-founder and CEO, while Antoine Carossio is identified as co-founder and CTO; public profiles and event materials describe Carossio as a former pentester for the French intelligence services and a former machine learning researcher at Apple, and one company profile says the founders met at UC Berkeley after seeing a lack of developer-oriented tooling for secure APIs [TechCrunch, June 2023] [Y Combinator] [Antoine Carossio | GraphQLConf 2023] [Startup Savant, October 2024].
Capital formation is one of the cleaner parts of the file. Publicly disclosed funding totals about $22.9 million across a reported $1 million pre-seed, a $3.9 million seed led by Iris in June 2023, and the $18 million Series A in March 2026 led by Balderton Capital, with participation from investors including Frst, Uncorrelated Ventures, and Y Combinator; the business model is SaaS [Startup Savant, October 2024] [TechCrunch, June 2023] [Tech.eu, March 2026].
Over the next 12 to 18 months, the practical questions are whether Escape can turn its broadened platform message into repeatable enterprise adoption, whether customer count claims above 100 translate into durable expansion revenue, and whether hiring into sales, customer success, and technical account management signals a functioning go-to-market motion rather than only product ambition [Built In, August 2026] [Work at a Startup, March 2026] [Y Combinator]. The public signal is encouraging but still incomplete: headcount appears to have grown from 10 at seed to roughly 29 to 30-plus by 2026, yet most operational claims beyond financing and hiring remain lightly corroborated [TechCrunch, June 2023] [Y Combinator] [Work at a Startup, March 2026].
Partially corroborated -- Funding history is corroborated across TechCrunch, Tech.eu, and Startup Savant, while product scope and customer traction rely partly on company materials and recruiting pages.
Taxonomy Snapshot
| Axis | Value |
|---|---|
| Stage | Series A |
| Business Model | SaaS |
| Industry / Vertical | Security |
| Technology Type | AI / Machine Learning |
| Geography | Western Europe |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (2) |
| Funding | $10M+ (total disclosed ~$22,900,000) |
Inside the Company
PUBLIC
Escape is a Paris-based cybersecurity startup founded in 2020, according to its Crunchbase profile [Crunchbase]. The company presents itself through its careers materials as an offensive security platform, with roots in API security testing and a broader push into automated application security workflows [Escape].
The public record on early company formation is thin, but the operating timeline is clearer from later financing and company materials. Escape appeared in Y Combinator as part of the W23 batch, according to its company profile there [Y Combinator], then surfaced more broadly with a June 2023 seed round that positioned the business around API vulnerability discovery [Crunchbase]. By March 2026, the company was describing a wider platform spanning offensive security automation in recruiting materials published under its own brand [Escape].
That sequence matters because it shows a company that seems to have expanded from a narrow API-testing wedge into a larger application security pitch without changing the underlying problem statement. Publicly available company materials do not establish the legal entity name in this section's allowed source set, so the cleanest confirmed facts are the 2020 founding date, Paris headquarters, Y Combinator participation, and subsequent financing milestones reflected on Crunchbase and company-controlled pages [Crunchbase] [Escape] [Y Combinator].
Partially corroborated -- Founded year and headquarters are supported by Crunchbase, while milestone detail in this section relies on company-controlled pages and platform profiles with partial corroboration.
Under the Hood
MIXED
Escape started with a narrower claim than its current positioning suggests. At seed, the company described a product that generated realistic API requests, simulated attacks with an AI-powered algorithm, identified flaws, and proposed remediations, with particular emphasis on business-logic issues in APIs rather than only conventional scanner findings [TechCrunch, June 2023]. By March 2026, public coverage and company recruiting materials framed the platform more broadly as an offensive-security system spanning business-logic testing, agentic pentesting, attack-surface management, code-to-cloud correlation, and DAST [Tech.eu, March 2026] [Escape].
That arc matters because it implies a move from a point solution in API security toward a wider application-security workflow. The public evidence still anchors most clearly in automated testing for APIs and modern applications, aimed at development and application-security teams before code reaches production [TechCrunch, June 2023]. Claims that the product can replace or supplement manual penetration testing and bug-bounty programs, and that AI agents can find and exploit vulnerabilities across applications, APIs, and microservices, come from job materials rather than independent product reviews, so they are better read as positioning than as externally validated capability [Escape, Unknown].
The technical picture is directionally clear even if the stack is not. Escape consistently presents the product as continuous offensive testing rather than static inspection, with AI agents acting against live systems to surface meaningful exploitable weaknesses [Escape, Unknown] [Tech.eu, March 2026]. The available public sources do not verify the underlying model architecture, deployment model, or integration depth; any stack assumptions would be speculative, aside from a general AI and security hiring focus (inferred from job postings) [Escape, Unknown].
Claim stands unchecked -- This section relies materially on company recruiting materials and press coverage that cites company descriptions, with limited independent product verification.
Market Research
PUBLIC
Security testing for modern applications matters now because software teams are shipping more API-driven and AI-assisted code, while the historical security stack was built for a narrower web application model [TechCrunch, June 2023] [Tech.eu, March 2026].
The evidence base here is narrower than investors would usually want for a full TAM view. No named third-party market sizing report was provided in the source set, so this section avoids unsupported TAM, SAM, or SOM figures and stays with observed demand signals around application security, API security, and automated offensive testing. What is clear from public reporting is that Escape started from a specific problem, API business-logic testing, because conventional scanners often missed the logic-layer issues that emerge in modern API architectures [TechCrunch, June 2023] [YesPress]. By March 2026, the company and covering press were positioning that wedge more broadly around AI cybersecurity agents and automation of offensive-security workflows, which suggests the relevant market is no longer just API testing but a larger application-security budget that also includes DAST, attack surface management, and pentest substitution [Tech.eu, March 2026] [Escape].
Demand drivers in the cited research are practical rather than theoretical. TechCrunch's 2023 coverage described Escape's early product as dynamically generating realistic requests, simulating attacks, and proposing remediations for APIs before production, which lines up with the need to secure software earlier in the development cycle [TechCrunch, June 2023]. The later recruiting materials point to a second tailwind, security teams looking to replace or supplement manual penetration testing and bug-bounty programs with continuous automated testing, especially across applications, APIs, and microservices [Escape] [Escape Careers]. That framing matters because it expands budget access beyond point-in-time testing and toward recurring platform spend, though the substitution claim remains company-sourced rather than independently benchmarked [Escape].
Adjacent and substitute markets are visible even without formal sizing data. Escape's current positioning overlaps with API security testing, DAST, attack surface management, and parts of the manual offensive-security services market, particularly where buyers want continuous validation instead of annual or event-driven testing [Escape Careers] [Escape]. The company is also leaning into AI-assisted software development as a source of new attack surface: one LinkedIn post attributed to an Escape employee said the firm's internal analysis identified more than 2,000 high-impact vulnerabilities across live applications built with automated coding tools, including 175 involving exposure of personal data [LinkedIn, retrieved 2026]. That claim is not independently verified, but it is directionally consistent with why this budget category is receiving attention.
Regulatory and macro pressure also support the category, even if the provided sources do not map them to a quantified market model. When teams ship customer-facing APIs and cloud applications, the cost of missed vulnerabilities increasingly includes data exposure risk, internal incident response burden, and the need to show repeatable security controls to customers during procurement [TechCrunch, June 2023] [Built In, August 2026]. The public materials also suggest a labor-market constraint: Escape was hiring across engineering, customer success, technical account management, and sales in 2026, which fits a market where buyers need tooling that compresses scarce security expertise into software workflows rather than relying only on specialist human testers [Work at a Startup, March 2026] [Y Combinator].
| Market lens | Public evidence | Relevance to Escape |
|---|---|---|
| API security testing | Escape began with API security testing focused on business-logic flaws [TechCrunch, June 2023] | Initial wedge and technical entry point |
| Automated offensive security | Escape aims to automate offensive-security work with AI agents [Tech.eu, March 2026] | Expands addressable budget beyond API scanning |
| DAST and attack surface management | Careers materials position the platform across DAST and attack surface management [Escape Careers] | Places the company in broader AppSec platform evaluations |
| Manual pentest and bug bounty substitution | Job materials say the product is intended to replace or supplement manual pentesting and bug bounty programs [Escape] | Indicates exposure to services and continuous testing budgets |
The table shows a market that is better understood as converging security budgets than as a single clean category. For investors, that can be attractive if the product lands through one pain point and expands, but it also means market definition will need validation through win-loss data rather than category labels alone.
Claim stands unchecked -- This section relies mainly on named press coverage and company recruiting or careers materials; no independent third-party market sizing report was provided, and several market-boundary claims remain company-sourced.
Competition and Substitutes
MIXED Escape is positioned less as a point API scanner and more as an attempt to collapse several offensive security workflows into one application-security platform, but the public record here is thinner on named rivals than on product ambition [TechCrunch, June 2023] [Escape] [Tech.eu, March 2026].
The competitive map is still legible even without source-grounded named peers. At the incumbent end, Escape is clearly selling against manual penetration testing and bug-bounty style workflows, which its own job materials frame as processes to replace or supplement with continuous automated testing [Escape]. In the adjacent tooling layer, its original wedge sat in API security and business-logic testing, an area where conventional scanners have often been weaker because they detect technical flaws more readily than misuse of application flows or GraphQL-specific logic [TechCrunch, June 2023] [YesPress]. The current product description broadens that pitch into attack surface management, DAST, code-to-cloud correlation, and agentic pentesting, which places Escape in competition not with one category but with buying bundles that security teams may otherwise source from several vendors [Escape] [Tech.eu, March 2026].
Escape's clearest edge today appears to be product scope anchored in a credible technical origin story. The company began with a narrow problem, testing API business logic before production, and has since expanded that thesis into a wider offensive-security platform rather than starting with a diffuse platform claim on day one [TechCrunch, June 2023] [Tech.eu, March 2026]. Founder-market fit also matters here: TechCrunch identifies Tristan Kalos as co-founder and CEO, while Antoine Carossio is identified as co-founder and CTO; separate public sources describe Carossio as a former pentester for French intelligence services and a former machine learning researcher at Apple [TechCrunch, June 2023] [Y Combinator] [Antoine Carossio | GraphQLConf 2023] [Startup Savant, October 2024]. That combination can help with buyer trust in security, but it is only durable if the company converts technical credibility into repeatable enterprise distribution. Public evidence for that conversion is suggestive, not decisive: recruiting pages cite more than 100 customers, while one employee LinkedIn profile claims 2,000+ customers, a gap large enough that the lower figure is the safer operating assumption [Built In, August 2026] [LinkedIn].
The exposure is straightforward. Because Escape now spans API security testing, DAST, attack surface management, and offensive automation, it risks being judged against specialists in each budget line while still needing to prove that one integrated platform is good enough to displace those specialists [Escape] [Tech.eu, March 2026]. Larger vendors and established application-security suites typically have stronger distribution, existing security operations relationships, and more procurement familiarity than a company that had 10 employees at seed in June 2023 and roughly 29 to more than 30 employees by 2026, depending on source [TechCrunch, June 2023] [Y Combinator] [Work at a Startup, March 2026]. The hiring mix also suggests the immediate constraint may be go-to-market reach rather than product surface alone: public listings include a Head of Sales, Customer Success Lead, Technical Account Manager, and multiple senior engineering roles, which is consistent with a company still building commercial coverage in parallel with the platform [Y Combinator] [Work at a Startup, March 2026] [Escape].
The most plausible 18-month scenario is a consolidation contest around whether buyers prefer a narrower best-of-breed tool or a broader automated offensive-security stack. Escape is a likely winner if security teams keep shifting budget toward continuous testing of production-like application flows and accept agentic tooling as a practical substitute for portions of manual pentesting, because that is where its public positioning is most coherent [Escape] [Tech.eu, March 2026]. Escape is more likely to lose ground if the category settles around vendors that already control the broader AppSec platform relationship, or if buyers treat automated offensive tooling as an add-on rather than a system of record, because the company does not yet appear, from public evidence, to own a dominant channel or a uniquely documented installed base [Built In, August 2026] [TechCrunch, June 2023].
Reasoned from indirect evidence -- Core positioning is corroborated by TechCrunch, Tech.eu,
Opportunity
PUBLIC
The prize here is unusually large if Escape can turn a narrow API-security wedge into a broader offensive-security control point inside modern software teams.
The most credible upside case is not that Escape becomes another point scanner, but that it becomes a default continuous testing layer for application security teams that now stitch together manual pentests, bug bounties, API testing, attack-surface management, and DAST workflows across separate vendors [TechCrunch, June 2023] [Escape] [Tech.eu, March 2026]. That outcome is reachable, not merely aspirational, because the public record already shows three things moving in the right direction: the company started with a concrete pain point in API business-logic testing [TechCrunch, June 2023]; it has broadened its product surface toward a fuller offensive-security workflow [Escape]; and investors have continued to fund that expansion through a reported $18 million Series A led by Balderton Capital in March 2026, after a $3.9 million seed in June 2023 [Tech.eu, March 2026] [TechCrunch, June 2023]. The customer count claim is still lightly evidenced, but recruiting materials that cite more than 100 customers, along with named customer references such as Zoom, Schibsted, Wiz, and PandaDoc, are directionally consistent with a company that has moved beyond an early technical experiment [Built In, August 2026] [Escape job posting].
A useful way to frame the upside is to separate product ambition from the concrete routes by which Escape could get there.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| API wedge becomes platform | Escape keeps winning on API and business-logic testing, then expands those accounts into DAST, attack-surface management, and broader agentic pentesting | The product scope described in recruiting and careers materials suggests a deliberate move from single use case to platform breadth [Escape] [Escape job posting] | Escape's earliest positioning focused on API business logic, a real blind spot in older tooling, while later materials show a broader control plane ambition [TechCrunch, June 2023] [Escape] |
| Standard tool for AI-built apps | Escape benefits from a rise in AI-assisted software development and positions itself as the testing layer for code that ships faster and breaks in less predictable ways | The Series A announcement explicitly ties the roadmap to AI cybersecurity agents, and company materials frame the product around code written and attacked at AI speed [Tech.eu, March 2026] [LinkedIn] | If application teams ship more code with less manual review, automated offensive testing becomes easier to justify in budget and workflow terms [Tech.eu, March 2026] [Escape job posting] |
| US enterprise distribution step-change | Escape's go-to-market matures from founder-led sales into a repeatable US commercial motion with sales, customer success, and technical account coverage | Open roles include Head of Sales, Customer Success Lead, and Technical Account Manager, while job listings place teams in New York, San Diego, Paris, and Amsterdam [Y Combinator] [Work at a Startup, March 2026] [Built In, August 2026] | The hiring pattern looks like a company building post-product commercial infrastructure rather than only adding engineers [Escape job posting] [Y Combinator] |
The compounding mechanism, if it develops, is a product-plus-distribution flywheel rather than a pure network effect. Escape's initial wedge was technical credibility in API and business-logic testing [TechCrunch, June 2023]. If that wedge lands security teams, the broader platform described in current materials could let the company sell adjacent modules into the same budget owner, while customer-facing hiring in sales, success, and technical account management should improve deployment speed, retention, and reference density over time [Escape] [Escape job posting] [Work at a Startup, March 2026].
There is also a smaller but potentially meaningful data advantage embedded in the story, although the public evidence is still company-led. Escape's current positioning emphasizes AI agents that discover and exploit vulnerabilities across applications, APIs, and microservices, and an internal analysis cited by an employee profile claims the company identified more than 2,000 high-impact vulnerabilities in live applications built with automated coding tools [Escape job posting] [LinkedIn]. If the platform is continuously testing real production-like environments across a growing customer base, the company could accumulate a proprietary corpus of exploit paths, remediation patterns, and false-positive suppression signals that improves product performance account by account. That is the kind of compounding that can make a security product harder to displace, though the degree of advantage is not independently verified in the available sources [LinkedIn] [Built In, August 2026].
The size of the win depends on whether Escape becomes a durable platform or remains a specialist feature. There is no cited market-size dataset in the available materials, so the cleaner public benchmark is funding and category formation rather than a formal TAM model. If the "API wedge becomes platform" scenario plays out, Escape could plausibly resemble a scaled application-security platform with meaningful strategic value to larger security buyers or late-stage investors, supported by the fact that the company has already progressed from a 10-person seed-stage team in 2023 to roughly 29 employees on Y Combinator's company page and more than 30 employees in a 2026 job listing, while raising a total disclosed $22.9 million across pre-seed, seed, and Series A rounds [TechCrunch, June 2023] [Y Combinator] [Work at a Startup, March 2026] [Startup Savant, October 2024] [Tech.eu, March 2026]. A concrete way to translate that into company value is this: if Escape becomes a credible multi-product application-security platform with a repeatable US enterprise motion, the outcome could support a venture-scale valuation in the high hundreds of millions of dollars, or more, in a future financing or strategic sale (scenario, not a forecast). The evidence does not yet support a tighter valuation band, but it does support the claim that this is a category with room for an independent winner if product breadth and commercial execution arrive together.
Claim stands unchecked -- This section relies on a mix of independent funding coverage from TechCrunch and Tech.eu, plus substantial company and employee-claimed product and customer evidence from Escape recruiting materials, Y Combinator listings, Built In, Work at a Startup, and LinkedIn.
Sources
Open sources
[Crunchbase] Escape | Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/escape-b74e
[Tech.eu, March 2026] Escape secures $18M Series A to develop AI cybersecurity agents | https://tech.eu/2026/03/10/escape-secures-18m-series-a-to-develop-ai-cybersecurity-agents/
[Startup Savant, October 2024] Escape | Startup Company Profile | TRUiC | https://startupsavant.com/startup-center/escape
[TechCrunch, June 2023] Escape dynamically scans APIs to find security flaws | https://techcrunch.com/2023/06/05/escape-dynamically-scans-apis-to-find-security-flaws/
[Y Combinator] Escape Jobs | https://www.ycombinator.com/companies/escape/jobs
[Built In, August 2026] Sales Director, US AI AppSec at Escape | https://builtin.com/job/sales-director-us-ai-appsec/10067864
[Work at a Startup, March 2026] Customer Success Lead, NYC at Escape | https://www.workatastartup.com/jobs/91356
[Escape] Escape Careers | https://escape.tech/careers
[Escape] Senior Full Stack Engineer | AI & Security | https://jobs.ashbyhq.com/escape/32e02dde-b428-4ed5-8a50-51e98fca5ddc
Articles about Escape
- Escape's AI Agents Are Penetration Testing for the API-First Era — The Parisian startup, now with over 100 customers, automates business-logic testing and attack simulation where legacy scanners fall short.