All analysis was generated autonomously, without human review. Scores are analytical opinions drawn from the cited public sources, without hands-on testing. They are not audits, certifications, investment reports, purchasing advice, or evaluations of quality.
NeuralTrust sells software that secures the AI agents large enterprises run, to banks, airlines, energy companies and government agencies. Its TrustGate gateway routes agents' calls to models and tools, TrustGuard blocks attacks in that traffic, and TrustTest simulates attacks on agents. Tech Funding News reports it was founded in 2022. It raised a $20 million seed round led by Alstin Capital. Its named customers are the banks Abanca and Banc Sabadell and the airlines Iberia and AirEuropa. NeuralTrust says Gartner named it a Pioneer among AI application security startups. Tech Funding News reports several of its government and banking customers seek a non-US vendor such as NeuralTrust. A customer routing its agents through TrustGate would have that wiring to redo before switching vendors.
| Description | NeuralTrust helps enterprises discover and secure the AI agents running across their teams, with products for runtime protection, a gateway that connects agents to models and tools, agent posture management, and adversarial red teaming. | [f1] |
|---|---|---|
| HQ | Barcelona, Spain | [f2] |
| Funding | $20M total | [f3] |
| Latest funding | Seed, $20M (June 2026, led by Alstin Capital) | [f3] |
| Deployment | Hybrid, SaaS, Self-hosted | [f4] |
| Product | What it does |
|---|---|
| NeuralTrust | NeuralTrust platform: TrustGate agent gateway, TrustGuard runtime security, TrustLens agent posture management, and TrustTest red teaming for LLMs and AI agents. |
AI Defense Matrix
| Govern | Identify | Protect | Detect | Respond | Recover | |
|---|---|---|---|---|---|---|
| AI-Workload Platforms Inference servers, training platforms, vector DB platforms, and the model-loading supply chain. | ||||||
| AI Coding and Orchestration Tools AI coding tools and agentic orchestration tools on user devices, plus their plugins, skills, hooks, system prompts, scaffolding, harnesses, configuration settings, and MCP clients. | ||||||
| AI-Generated Code Code produced by AI tools, AI-assisted reviews, AI-generated infrastructure-as-code and tests, and vibe-coded apps that bypass CI/CD. | ||||||
| AI Traffic Paths to external and self-hosted AI services, MCP tool channels, agent-to-agent calls, model-registry downloads, and egress to unapproved AI services. AI gateways, LLM routers, and MCP gateways steer traffic along those paths. The steering decisions and their inputs belong here too: routing policies, MCP server registries, and agent naming and discovery services. | ||||||
| AI Model Model weights, fine-tuning checkpoints, model cards, registries, an AI bill of materials (AIBOM), and the third-party LLMs your enterprise consumes. | ||||||
| Training Data Datasets used for training, fine-tuning, and continued learning. | ||||||
| Runtime AI Data User prompts, inference inputs, RAG content, vector DB content, persistent agent memory, and interaction history. | ||||||
| AI Agent Identities AI agents as non-human principals, plus credentials, keys, permission scopes, service accounts, and delegation chains across agents and tools. |
NeuralTrust is an AI gateway and runtime firewall that screens LLM and agent traffic for injection attacks and data leaks, plus automated red teaming. It is mapped to the AI Defense Matrix. [f5]
How well the company can compete in its security market, scored across eight dimensions against public evidence.
| Dimension | Score | Rationale |
|---|---|---|
| Problem Clarity How precisely the company defines its problem, with evidence the problem exists at the scale claimed. | 3/5 | NeuralTrust names a clear buyer, the security team at a large enterprise whose AI agents run across teams, model providers, and vendor software, and The Next Web restates that sprawl problem in its own words. The cited figures are a Gartner category projection and NeuralTrust's own interaction count, without independent quantification for its buyers, so the score is 3. [s8, s20, s25] |
| Capability Depth How specific the technical capabilities are, with evidence beyond marketing claims such as docs and third-party validation. | 4/5 | NeuralTrust's product pages detail TrustGuard's multi-turn detection and TrustGate's routing of model, MCP, and agent-to-agent calls, and TrustGate itself is published on GitHub as open-source Go code anyone can inspect. The Echo Chamber jailbreak research is an arXiv preprint with a co-author from ICREA and UPF, and BankInfoSecurity reports NeuralTrust's account of applying mitigations to its own gateway. The inspectable gateway code and the co-authored preprint are the two external validation points that keep the score at 4. [s2, s3, s17, s23, s21, s24] |
| Market Timing Whether the market is ready for this product, with evidence that buyers are actively seeking solutions. | 3/5 | AI agents that companies deploy faster than they can track them are a credible enabler, and in June 2026 The Next Web reported European institutions swapping US security vendors for home-grown ones as the EU AI Act tightens. That is one independent buyer-side kind. The Gartner Pioneer placement reaches the record through NeuralTrust's own announcement, and the doubled annual recurring revenue is the company's figure. The Next Web also reports the lead investor conceding that timing is the open question, whether companies buy before a serious agent incident. [s25, s7, s8, s20] |
| Team Credibility Demonstrated domain expertise with public signals such as prior exits, publications, and industry recognition. | 3/5 | Tech Funding News reports that CEO Joan Vendrell was chief data officer at Mango after Amazon and McKinsey and that CTO Victor Garcia designed security systems for telecoms and banks, verifiable senior experience with a direct security thread. The Echo Chamber preprint and its press coverage add one research event, and the exit the company credits to COO Alejandro Domingo names no security product, so the plurality rung 4 needs is not evidenced and the score is 3. [s20, s8, s23, s21] |
| GTM Proof Evidence of actual traction (customers, revenue signals, partnerships) beyond stated intentions. | 4/5 | Tech Funding News lists Iberia, AirEuropa, Abanca and Banc Sabadell as customers in its own voice, and the AppSec Santa review names Banco Sabadell, Iberia, Capgemini, NTT Data, Air Europa, ABANCA and ISDIN. NeuralTrust names the same four and KPMG as a strategic partner, and an ABANCA transformation director describes stress-testing a chatbot with it. That is multiple named references, multiply sourced. Scale stays unconfirmed, since the billion-dollar-customer share and the doubled annual recurring revenue are the company's figures, with no amounts. [s20, s26, s8, s1] |
| Funding Efficiency Whether funding matches go-to-market ambition, with signs of capital-efficient growth. | 2/5 | NeuralTrust raised a $20 million seed led by Alstin Capital and describes it as a record for an EU cybersecurity seed, a claim The Next Web called worth a pinch of salt but not outlandish. Against a raise of that size, the verifiable commercial proof is four named customers and a doubling of annual recurring revenue with no disclosed figures, so the raise outruns verifiable results and the score is 2. [s8, s20, s25] |
| Category Clarity Whether the company creates or fits a recognizable category that buyers can quickly place in their stack. | 3/5 | NeuralTrust reports Gartner placements in AI gateways, guardian agents, AI application security, and AI runtime defense, plus KuppingerCole and MarketsandMarkets leader placements, so analysts do place its products. The category labels differ by analyst, and NeuralTrust's own account of the Gartner quadrant describes a nascent market whose competitive landscape is still forming, which holds the score at 3. [s7, s8, s11, s9] |
| Incumbent Defensibility How vulnerable the core value proposition is to absorption as a feature by a platform vendor. | 3/5 | TrustGate routes the agent calls a customer sends through it to one controlled point and TrustGuard tracks attacks across conversation turns, which gives a buyer some integration friction to unwind before bundling a substitute. Tech Funding News and The Next Web both report Palo Alto Networks buying Protect AI, and the record documents no data flywheel or procurement lock that a platform vendor could not eventually match, so the score is 3. [s2, s3, s20, s25] |
NeuralTrust sells to the security team at a large enterprise that has lost track of its AI agents. The company describes one team building a customer-service agent on one model provider while another automates back-office work on a different one. The Next Web restates the problem as agent sprawl, with agents multiplying inside large companies without anyone keeping a tally.
The buyer's pain is agents that act on internal systems without oversight. NeuralTrust says it inspects millions of agent interactions a day and that around 1.2% are malicious, attempts to extract data, hijack a tool, or push an agent past its rules. That figure is the company's own measurement, and Tech Funding News relays a Gartner projection that 40% of enterprises will shut down AI agents after finding governance problems in production. [s8, s20, s25, s1]
NeuralTrust sells four products that cover an agent from discovery to runtime. TrustLens builds an inventory of agents and maps each agent, model, tool, data source, and identity as a connected graph. TrustGate is the gateway for the model, MCP server, tool, and agent-to-agent calls a customer routes through it, and TrustGuard inspects agent traffic it collects from gateways, SDKs, browsers, and log streams.
TrustGuard's detection follows context across a whole conversation. The company says it catches multi-turn attacks where a jailbreak fails once and succeeds on a later attempt, and it can allow, block, transform, or only alert depending on detection fidelity. TrustTest is the offensive side, a code-first red-teaming framework that runs attacks against a customer's models and agents and grades the results.
The products connect in a loop the company describes directly. NeuralTrust says TrustTest findings feed TrustGuard policy configuration, joining testing to runtime enforcement. TrustGate is also published on GitHub as open-source Go code covering multi-provider routing, guardrails, semantic caching, and MCP support. [s1, s2, s3, s4, s5, s7, s14, s17]
NeuralTrust competes with other AI security specialists and with platform vendors buying their way into the category. An AppSec Santa review compares it with Lakera Guard and notes that Check Point acquired Lakera. Tech Funding News reports that Palo Alto Networks bought Protect AI for about $500 million in April 2025, which shows a large security vendor can own agent security through acquisition.
NeuralTrust positions itself as the European option. Its CEO told Tech Funding News that the main competitors come from the United States and that some customers, especially governments, care where their technology comes from. The Next Web describes that pitch as well timed, citing the tightening EU AI Act and European institutions replacing US security vendors with home-grown ones. [s26, s20, s25]
NeuralTrust names AirEuropa, Abanca, Iberia, and Banc Sabadell among its customers, and Tech Funding News lists the same four. The company says 92% of its customers report more than $1 billion in annual revenue and 80% are based in Europe. On its own site, ABANCA's director of transformation says NeuralTrust helped stress-test the bank's generative AI chatbot before launch.
The company reports revenue growth without disclosing figures. NeuralTrust says it doubled its full-year 2025 annual recurring revenue during Q1 2026, and Tech Funding News notes that it did not share exact numbers. NeuralTrust names KPMG as a strategic partner for large enterprise clients, alongside collaborators including Capgemini and Sopra Steria.
Analyst recognition has accumulated through 2026, as reported by the company. NeuralTrust announced a Representative Vendor listing in Gartner's Market Guide for AI Gateways in October 2025 and a Pioneer placement in Gartner's new Emerging Market Quadrant for AI application security in September 2026. It also reports Sample Vendor listings in the 2026 Gartner Hype Cycles for Infrastructure Security and for Data Security Technologies and opened a London office at 167 Great Portland Street in September 2026. [s8, s20, s1, s7, s9, s11]
NeuralTrust's founders combine enterprise data, consulting, and security architecture experience. Tech Funding News reports that CEO Joan Vendrell was chief data officer at Mango and worked at Amazon and McKinsey, that Alejandro Domingo led AI transformation projects as a McKinsey associate partner, and that Victor Garcia designed security systems for telecoms and banks. The company adds that Domingo previously scaled a startup to an exit.
The research team gives the company a public technical record. NeuralTrust researcher Ahmad Alobaid developed the Echo Chamber multi-turn jailbreak, which The Hacker News and BankInfoSecurity covered and which appears as an arXiv preprint co-authored with a researcher from ICREA and UPF. NeuralTrust says it contributed to the OWASP GenAI Red Teaming Manual, where Echo Chamber is featured as a real-world example. [s20, s8, s23, s21, s24, s9]
NeuralTrust holds an ISO/IEC 27001:2022 certification that a procurement team can check. The company says Insight Assurance issued it after an independent audit, that it covers the systems behind its cloud products and customer data processing, and that the certificate runs through 2029 with annual surveillance audits.
NeuralTrust built its deployment options to keep sensitive prompts inside the customer's environment. The company separates a control plane from a data plane, and in its hybrid and on-premises modes the data plane runs inside the customer's own cloud or data center. It offers SaaS, a hybrid mode with a privately hosted data plane, and an on-premises or air-gapped mode aimed at defence, government, and financial-sector buyers. [s10, s1, s2]
| Company | Relationship | Note | Compare |
|---|---|---|---|
| Lakera | competes with | An AppSec Santa review compares NeuralTrust with Lakera Guard for runtime prompt-injection defense, and the review describes NeuralTrust as EU-headquartered and Lakera as European. | |
| Prompt Security | competes with | Competes for the same enterprise buyer securing employee and homegrown AI use at runtime. | |
| Aim Security | competes with | Competes for the enterprise security team that needs runtime controls over LLM and agent traffic. | |
| Lasso Security | competes with | Competes for the same enterprise AI security buyer as NeuralTrust. | |
| Palo Alto Networks | adjacent | Tech Funding News reports it bought Protect AI, which puts a large security platform vendor in the AI security budget NeuralTrust sells into. | N/AWe scored these companies at different scopes, so the totals measure different things. |
| Cloudflare | adjacent | A network platform vendor adjacent to NeuralTrust's AI gateway for the same enterprise traffic-control budget. | |
| OpenAI | adjacent | A model provider whose models NeuralTrust's TrustGate gateway routes agent calls to. | N/AWe scored these companies at different scopes, so the totals measure different things. |
Add analyzed competitors to compare them side by side with NeuralTrust.
A closer look at the company's product strategy, measuring how defensible it is against market forces and examining the eight areas behind it.
reinforce or reposition
NeuralTrust sells software that secures enterprise AI agents. Its TrustGate gateway brokers the calls to AI models and tools a customer routes through it, and TrustGuard enforces policy on that agent traffic. NeuralTrust names banks and airlines among its customers, including Abanca, Banc Sabadell, Iberia, and AirEuropa. Tech Funding News reports it was founded in 2022. It raised a $20 million seed led by Alstin Capital. NeuralTrust reports a Pioneer placement in Gartner's Emerging Market Quadrant for AI application security. It publishes its attack research, including the Echo Chamber jailbreak, a multi-turn attack that defeats AI models' safety mechanisms. A customer that wired its agents into TrustGate and TrustGuard would have that integration work to redo before switching vendors.
| Dimension | Score | Rationale |
|---|---|---|
| Value Delivery Does the product sell software as the product, or judgment, trust, or accountability with software as the delivery mechanism. | 1/3 | NeuralTrust sells TrustGate, TrustGuard, TrustLens, and TrustTest as software in SaaS, hybrid, and on-premises modes, priced by protected apps and agents, traffic, and deployment model. |
| Switching Cost How expensive leaving is for a customer: data portability, integrations, learned workflows, network effects, regulatory data residency. | 2/3 | Routing agents' model, MCP, and tool calls through TrustGate and wiring TrustGuard collectors into gateways and SDKs creates real integration friction. The cited sources do not quantify the cost of re-architecting those flows, so the score is 2. |
| Compliance Moat Whether certifications, liability acceptance, or audit trails block an easy replacement. | 1/3 | NeuralTrust holds ISO/IEC 27001:2022, an attestation a funded competitor can also obtain. Its air-gapped mode lists a FedRAMP and IL4 pathway, a route toward authorization that a rival can also pursue, so no requirement in the record blocks replacement. |
| Problem Complexity Whether the product requires ML, optimization, real-time systems, or years of specialized expertise. | 3/3 | NeuralTrust runs a gateway it rates at more than 20,000 requests per second per node, tracks multi-turn attacks across conversation turns, and runs automated red teaming. Its researchers published the Echo Chamber jailbreak as an arXiv preprint, so the work combines real-time systems with adversarial research. |
| Buyer Profile Whether buyers are SMB operators, mid-market IT teams, or regulated enterprises and governments with procurement gates. | 3/3 | NeuralTrust names banks and airlines as customers, Abanca, Banc Sabadell, Iberia, and AirEuropa, and says its customers include government agencies, so its buyers are regulated enterprises and governments. |
| Layer Whether the product is an end-user application, a platform with application features, or infrastructure other applications depend on. | 2/3 | TrustGate is the gateway in front of models, tools, and MCP servers. NeuralTrust also sells TrustLens posture views and TrustTest red-teaming runs, and its documentation says customers can adopt TrustTest separately from TrustGate and TrustGuard as part of a platform with application features. |
| Proprietary Data, Content, or IP Whether the product accumulates datasets, content licenses, or IP that a rival cannot recreate from scratch. | 1/3 | NeuralTrust publishes its attack research, releasing Echo Chamber as an arXiv preprint with open code on GitHub. TrustTest's built-in probe library is described without its sources or any statement that it is kept private, so the record evidences no retained dataset, licence, or patent. |
NeuralTrust sells to large, regulated enterprises that run AI agents across many teams. It names airlines and banks as customers, AirEuropa, Iberia, Abanca, and Banc Sabadell, and says its customers also include energy companies and government agencies. By the company's count, 92% of its customers report more than $1 billion in annual revenue.
The customer base is concentrated in Europe, and the company treats that as a selling point. NeuralTrust says 80% of its customers are based in Europe. Tech Funding News reports that several of its government and banking customers are seeking providers headquartered outside the United States as EU AI Act requirements tighten.
The deployment options match that buyer. NeuralTrust offers an on-premises or air-gapped mode it aims at defence, government, and the financial sector, and a hybrid mode that keeps the data plane inside the customer's boundary. It opened a London office in September 2026 and tells North American prospects it supports customers in US time zones.
NeuralTrust's central product is a gateway for the agent calls a customer routes through it. TrustGate routes calls to OpenAI, Anthropic, or self-hosted models, scopes each agent's tool access per user, and traces agent-to-agent handoffs, with inline prompt inspection, PII redaction, and semantic caching. The company publishes the gateway on GitHub as open-source Go code.
TrustGuard adds detection that follows a whole conversation. The company says it tracks context across turns and models interaction patterns over time, flagging probing and slow escalation before an attack completes. TrustLens inventories agents as a connected graph of models, tools, data sources, and identities, and TrustTest runs automated red-teaming attacks and grades the target's responses.
BankInfoSecurity reports that NeuralTrust applied Echo Chamber mitigations to its gateway. NeuralTrust researcher Ahmad Alobaid developed the Echo Chamber multi-turn jailbreak. The Hacker News reported that the attack succeeded more than 90% of the time on several harmful-content topics in a controlled test against OpenAI and Google models. BankInfoSecurity reports the company disclosed the findings to both model makers and applied mitigations to its own gateway, and the work appears as an arXiv preprint.
NeuralTrust sells directly to large enterprises and through consulting partners. It names KPMG as a strategic partner for large enterprise clients and Capgemini and Sopra Steria among its collaborators. A free TrustGate sign-up lets a team route traffic without a sales call, while the full platform is sold through demos and tailored quotes.
The company names four customers and reports growth without figures. NeuralTrust and Tech Funding News both list AirEuropa, Abanca, Iberia, and Banc Sabadell, and the company says it doubled its full-year 2025 annual recurring revenue during Q1 2026.
Analyst placements are a visible part of how the company presents itself. NeuralTrust reports a Representative Vendor listing in Gartner's Market Guide for AI Gateways in October 2025 and a Pioneer placement in Gartner's new Emerging Market Quadrant for AI application security in September 2026. It raised a $20 million seed to hire engineers, integrate its products, and expand across Europe, and Tech Funding News reports plans to enter the US market.
NeuralTrust prices by deployment size rather than publishing a rate card. Its homepage says enterprise pricing depends on protected apps and agents, traffic, and deployment model, with a tailored quote after a short discovery or proof of value. An AppSec Santa review confirms that NeuralTrust does not publish dollar amounts.
A free entry point sits in front of the enterprise sale. NeuralTrust lets teams sign up free and start using TrustGate in minutes without a sales call, then upgrade as governance and compliance needs grow.
NeuralTrust delivers software in three deployment modes. The SaaS mode comes with managed infrastructure and automatic upgrades, the hybrid mode keeps a SaaS control plane with a privately hosted data plane, and the on-premises or air-gapped mode makes no external calls. The company says the control plane never processes customer content.
The runtime components enforce decisions inline. TrustGuard collectors ingest agent traffic from gateways, SDKs, browsers, sidecars, and log streams, and TrustGuard allows, blocks, or transforms requests in flight, or only alerts, depending on detection fidelity. The company publishes a throughput figure of more than 20,000 requests per second per node.
The cited sources describe software delivery and do not document a managed security service. The documentation covers securing coding agents such as Claude Code and Cursor, and NeuralTrust's GitHub publishes a Helm chart for the platform.
NeuralTrust holds its own ISO/IEC 27001:2022 certification, issued by Insight Assurance after an independent audit. The company says the certificate covers the systems behind its cloud products and customer data processing and runs through 2029 with annual surveillance audits. A buyer's security review can start from that certificate.
NeuralTrust designed its deployment modes around where sensitive prompts go. The company says its data plane runs inside the customer's cloud or data center and enforces policy locally, and that it never trains on customer data. For its air-gapped mode, the company lists a FedRAMP and IL4 pathway, a route toward those federal authorizations.
NeuralTrust positions its gateway as the control point for all enterprise agent traffic. TrustGate brokers calls to models, MCP servers, tools, and other agents, and the documentation covers OpenAI, Anthropic, Gemini, Azure OpenAI, Bedrock, and Vertex AI as providers. NeuralTrust describes its infrastructure-level security as a centralized gateway that applies one policy across all apps and models.
NeuralTrust publishes integrations for the tools agents already run in. The documentation covers securing Claude Code, Cursor, Codex, Windsurf, and GitHub Copilot, and NeuralTrust publishes integrations on GitHub for frameworks such as LangChain and Haystack. TrustGate, TrustGuard, and TrustTest share organization settings, and customers can adopt one without the others.
NeuralTrust's founders come from enterprise data, consulting, and security architecture. Tech Funding News reports that CEO Joan Vendrell was chief data officer at Mango after Amazon and McKinsey, that Alejandro Domingo was a McKinsey associate partner, and that Victor Garcia designed security systems for telecoms and banks.
The research team publishes attack techniques the products defend against. Researcher Ahmad Alobaid developed the Echo Chamber jailbreak, covered by The Hacker News and BankInfoSecurity and written up as an arXiv preprint. NeuralTrust says Echo Chamber and its Semantic Chaining attack are part of the OWASP AI Security Project taxonomy and that it contributed to the OWASP GenAI Red Teaming Manual.
| Id | Source | Tier | Accessed |
|---|---|---|---|
| f1 | NeuralTrust: The Platform for AI and Agent Security | official | 2026-09-22 |
| f2 | About us | NeuralTrust | official | 2026-06-14 |
| f3 | Tech.eu: NeuralTrust closes 20M to expand AI agent security platform | press | 2026-06-17 |
| f4 | AI Defense Matrix Catalog entry | other | 2026-06-10 |
| f5 | AI Defense Matrix Catalog mapping | other | 2026-06-23 |
This site is an experimental research aid created by Zeltser Security Corp. All its data gathering and analysis was performed autonomously without human review, and it can contain errors of fact, interpretation, and judgment that a human reviewer might catch.
The analyses are statements of opinion, not statements of fact. Machine analysis produced the scores, summaries, and matrix placements by weighing the public sources each page cites, and reasonable people can weigh the same sources differently. Where a page states a fact, it cites the public source and the date it was checked, and the statement is only as accurate as that source. Unless a profile expressly says otherwise, the analysis involves no hands-on testing and no independent validation of any company's products or services.
Nothing here is professional, security, legal, financial, investment, or purchasing advice, and nothing here is a recommendation to invest in, do business with, or avoid any company. Inclusion of a company is not an endorsement, and absence of a company is not a judgment about it. Reading this site creates no advisory or client relationship. Verify any detail you plan to act on against the vendor's current materials.
The content is provided "as is" and "as available," with all warranties disclaimed, express or implied, including merchantability, fitness for a particular purpose, accuracy, and non-infringement. No entry is warranted to be complete, current, or correct. Companies change, vendors update their claims, sources can be wrong, and automated analysis can misread them.
To the fullest extent permitted by law, the operator, Zeltser Security Corp, is not liable for any damages that arise from using this site or relying on its content, including direct, indirect, incidental, special, and consequential damages and lost profits, even if advised that such damages were possible. If you are dissatisfied with the site or disagree with these terms, your remedy is to stop using it.
Entries link to vendor pages, press coverage, and other external sites that Zeltser Security Corp does not control and is not responsible for. A link is not an affiliation with the destination or an endorsement of it. Product and company names and trademarks are the property of their owners, used here nominatively to identify the companies described. Short quotations from cited sources appear for identification and commentary.
Use, quotation, automated retrieval, and redistribution of the content are governed by the Terms of Use at cybercompanyprofiles.com/terms, which permit personal and internal business use with attribution and prohibit republication and resale.