TrojAI by A10 is the acquired AI-security product line originating from Saint John-based TrojAI Inc. A10 currently presents runtime protection, model testing and security controls for AI applications, agents and Model Context Protocol deployments.
Operating context
TrojAI operates within A10's AI-security portfolio, offering model testing, runtime protection, agent discovery and MCP controls through customer-controlled cloud or on-premises deployments. A10 describes discovery of MCP servers and tools, policy enforcement across agent communications and inspection of agent execution traces; these are marketed functions rather than proof of Canadian defence adoption.
The Wiz integration was publicly announced by TrojAI on March 31, 2026 and described again by A10 on September 8. Wiz discovers AI models and endpoints, TrojAI Detect uses that inventory for security testing, findings return to Wiz for contextual review, and TrojAI Defend supplies runtime enforcement. Wiz's own integration marketplace corroborates this workflow, while A10 offers marketplace access and guided onboarding. The September article therefore documents an existing integration rather than establishing its launch date. Separately, ISED lists the earlier T7-018032 AI/ML risk-management innovation for additional sales through June 21, 2027; this status should not be extended to every subsequent integration.
Canadian footprint
The Canadian Trademarks Journal identifies TrojAI Inc. at 14 King Street, Suite 102 in Saint John, New Brunswick. This public address does not establish current Canadian headcount, facility functions or a separate post-acquisition sales office.
Saint John, New Brunswick is the published city-level context for TrojAI. It does not imply a street address or exact facility location.
TrojAI's Wiz integration gives buyers an identifiable workflow from AI asset inventory to testing, contextual findings and runtime controls. Deployment architecture, data movement and procurement coverage still require configuration-specific confirmation.
Interpretation of the public record. Not qualification, endorsement or evidence of buyer interest.
Protect and evaluate AI applications, models, agents and Model Context Protocol deployments with TrojAI Defend, TrojAI Detect and Defend for MCP, including customer-controlled cloud or on-premises deployment options.
Evidence of maturity
TrojAI is a current A10 AI-security offering after the completed June 2026 acquisition. A10 announced AI Gateway general availability on August 13 and describes it as complementing TrojAI across the AI lifecycle; independent deployment-specific performance remains undisclosed.
Commercial availability
TrojAI is offered through A10/TrojAI sales and demonstration routes, with customer-controlled deployment options. Product packaging, supported integrations and Canadian contracting arrangements are agreed per engagement.
Core features
Runtime prompt-injection, jailbreak and data-leak protection
AI model security testing and evaluation
Agent and Model Context Protocol discovery and control
Customer-controlled cloud or on-premises deployment
Mappings to OWASP, MITRE and NIST AI-security guidance
Agent-action, MCP invocation and tool-call telemetry alongside coding-agent protection
Recorded applications
Securing generative-AI applications
Testing models before deployment
Governing enterprise AI agents and MCP connections
Protecting sensitive AI workflows in controlled environments
On August 13, 2026, A10 Networks launched A10 AI Gateway and described TrojAI by A10 security as part of the centralized control plane for AI operations.
Commercially presented by A10 Networks as TrojAI by A10, including TrojAI Defend, TrojAI Detect and Defend for MCP; exact packaging, pricing, supported deployment scale and Canadian contracting entity require direct confirmation.
Ownership and entity scope
A10 Networks acquired TrojAI Inc. on June 15, 2026 and now markets the technology as TrojAI by A10; transaction terms, ownership percentage and the precise post-acquisition subsidiary structure were not disclosed.
Reviewed connections
Mission area · Reviewed connectionEvidence strength: Needs Review
AI model, agent and application security could help protect autonomous monitoring software that uses those components, but no public source reviewed here establishes a specific patrol-system integration or deployment.
Reviewed connections indicate possible relevance based on public evidence. They do not indicate procurement direction, eligibility, endorsement or customer interest.
Questions for a first conversation
01
Which Wiz and TrojAI components, data exchanges and licensing arrangements are required for the proposed deployment, and are they covered by the intended procurement route?
The integration exchanges model inventory and security findings between two platforms. ISED's additional-sales record identifies an earlier named innovation, so deployment boundaries and procurement coverage for the new integration should be validated rather than inferred.
Original sources 10 linked records
These records support the organization, its owned capabilities and their reviewed connections. A source count is not a count of independent confirmations.
Wiz's marketplace describes discovery of models and endpoints, inventory transfer to TrojAI Detect for automated testing, security findings returned to the Wiz graph, and TrojAI Defend runtime policy enforcement.
Wiz's marketplace describes discovery of models and endpoints, inventory transfer to TrojAI Detect for automated testing, security findings returned to the Wiz graph, and TrojAI Defend runtime policy enforcement.
Wiz's marketplace describes discovery of models and endpoints, inventory transfer to TrojAI Detect for automated testing, security findings returned to the Wiz graph, and TrojAI Defend runtime policy enforcement.
Wiz's marketplace describes discovery of models and endpoints, inventory transfer to TrojAI Detect for automated testing, security findings returned to the Wiz graph, and TrojAI Defend runtime policy enforcement.
Supports: Profile · operating context
Wiz's marketplace describes discovery of models and endpoints, inventory transfer to TrojAI Detect for automated testing, security findings returned to the Wiz graph, and TrojAI Defend runtime policy enforcement.
Wiz's marketplace describes discovery of models and endpoints, inventory transfer to TrojAI Detect for automated testing, security findings returned to the Wiz graph, and TrojAI Defend runtime policy enforcement.
True North Map derives Canadian-operation, deployment, threat-validation and procurement-route questions from the acquisition, product and ISED evidence.
True North Map derives Canadian-operation, deployment, threat-validation and procurement-route questions from the acquisition, product and ISED evidence.
A10 currently presents three TrojAI by A10 product components but does not publish complete pricing or Canadian contracting terms.
Supports: Profile · operating context
A10 presents TrojAI within its portfolio with self-contained on-premises or customer-cloud deployment, model testing, runtime defence, agent/MCP-server discovery, agent trace analysis, MCP tool/server policy enforcement and malicious-traffic blocking. These product assertions do not evidence Canadian defence adoption.
The acquired identity, current products, dated integration, Canadian legal footprint, limits and reviewed questions support human review for the full editorial template.
Public sources cited · Facts and assessments kept separate · Human review
ISED lists TrojAI Inc. proposal T7-018032, Risk Management and Security of AI/ML Models, as eligible for additional sales until June 21, 2027. This listing identifies the tested innovation, not an award or automatic eligibility for the new Wiz integration.
True North Map derives Canadian-operation, deployment, threat-validation and procurement-route questions from the acquisition, product and ISED evidence.
Supports: Profile · operating context
ISED lists TrojAI Inc. proposal T7-018032, Risk Management and Security of AI/ML Models, as eligible for additional sales until June 21, 2027. This listing identifies the tested innovation, not an award or automatic eligibility for the new Wiz integration.
Supports: Profile · executive relevance summary
Source location: executiveRelevanceSummary
True North Map assesses the current product as decision-relevant for AI-security validation while preserving deployment, performance and Canadian operating unknowns.
True North Map derives Canadian-operation, deployment, threat-validation and procurement-route questions from the acquisition, product and ISED evidence.
True North Map derives Canadian-operation, deployment, threat-validation and procurement-route questions from the acquisition, product and ISED evidence.
ISED lists TrojAI Inc. proposal T7-018032, Risk Management and Security of AI/ML Models, as eligible for additional sales until June 21, 2027. This listing identifies the tested innovation, not an award or automatic eligibility for the new Wiz integration.
True North Map derives Canadian-operation, deployment, threat-validation and procurement-route questions from the acquisition, product and ISED evidence.
ISED lists TrojAI Inc. proposal T7-018032, Risk Management and Security of AI/ML Models, as eligible for additional sales until June 21, 2027. This listing identifies the tested innovation, not an award or automatic eligibility for the new Wiz integration.
True North Map derives Canadian-operation, deployment, threat-validation and procurement-route questions from the acquisition, product and ISED evidence.
ISED lists TrojAI Inc. proposal T7-018032, Risk Management and Security of AI/ML Models, as eligible for additional sales until June 21, 2027. This listing identifies the tested innovation, not an award or automatic eligibility for the new Wiz integration.
True North Map derives Canadian-operation, deployment, threat-validation and procurement-route questions from the acquisition, product and ISED evidence.
ISED lists TrojAI Inc. proposal T7-018032, Risk Management and Security of AI/ML Models, as eligible for additional sales until June 21, 2027. This listing identifies the tested innovation, not an award or automatic eligibility for the new Wiz integration.
A10's current product catalogue describes runtime AI protection, model evaluation and agent or MCP security under the TrojAI by A10 name.
The current TrojAI by A10 product route directs inquiries through A10 Networks' corporate contact channel.
Supports: Profile · operating context
A10 describes customer-controlled deployment, AI runtime protection, testing, agent discovery and MCP controls without naming Canadian defence adoption.
Supports: TrojAI by A10 AI security platform · core features
A10's current TrojAI catalogue supports runtime protection, model security testing, agent and MCP discovery or control, customer-controlled deployment and common AI-security framework mappings.
Supports: TrojAI by A10 AI security platform · defence applications
A10's current TrojAI catalogue supports runtime protection, model security testing, agent and MCP discovery or control, customer-controlled deployment and common AI-security framework mappings.
Supports: TrojAI by A10 AI security platform · name
A10's current TrojAI catalogue supports runtime protection, model security testing, agent and MCP discovery or control, customer-controlled deployment and common AI-security framework mappings.
Supports: TrojAI by A10 AI security platform · technical tags
A10's current TrojAI catalogue supports runtime protection, model security testing, agent and MCP discovery or control, customer-controlled deployment and common AI-security framework mappings.
Supports: TrojAI by A10 AI security platform · commercial availability
A10's current TrojAI catalogue supports runtime protection, model security testing, agent and MCP discovery or control, customer-controlled deployment and common AI-security framework mappings.
Supports: TrojAI by A10 AI security platform · maturity
A10's current TrojAI catalogue supports runtime protection, model security testing, agent and MCP discovery or control, customer-controlled deployment and common AI-security framework mappings.
Supports: TrojAI by A10 AI security platform · capability type
A10's current TrojAI catalogue supports runtime protection, model security testing, agent and MCP discovery or control, customer-controlled deployment and common AI-security framework mappings.
Supports: TrojAI by A10 AI security platform · summary
A10's current TrojAI catalogue supports runtime protection, model security testing, agent and MCP discovery or control, customer-controlled deployment and common AI-security framework mappings.
Supports: TrojAI by A10 AI security platform · technicaldomainslugs
True North Map classifies the current product under communications and cyber and mission software and data. The retained autonomous-patrol-and-monitoring link is rewritten as a needs-review software-security assessment, not a deployment or adoption claim.
Supports: TrojAI by A10 AI security platform → Autonomous Patrol and Monitoring
True North Map classifies the current product under communications and cyber and mission software and data. The retained autonomous-patrol-and-monitoring link is rewritten as a needs-review software-security assessment, not a deployment or adoption claim.
Supports: TrojAI by A10 AI security platform → Autonomous Patrol and Monitoring
True North Map classifies the current product under communications and cyber and mission software and data. The retained autonomous-patrol-and-monitoring link is rewritten as a needs-review software-security assessment, not a deployment or adoption claim.
A10 announced AI Gateway general availability on August 13 2026, describing it as complementing TrojAI and ThreatX across testing, runtime and network/API security. Software or integrated hardware runs in customer-controlled on-premises/private-cloud/air-gapped environments; this does not prove every product is a single integrated configuration.
Supports: TrojAI by A10 AI security platform · maturity
A10 announced AI Gateway general availability on August 13 2026, describing it as complementing TrojAI and ThreatX across testing, runtime and network/API security. Software or integrated hardware runs in customer-controlled on-premises/private-cloud/air-gapped environments; this does not prove every product is a single integrated configuration.
On September 8, 2026 A10 described TrojAI integration with Wiz AI-APP, connecting model and endpoint discovery with testing and runtime protection. TrojAI had already announced the integration on March 31, 2026, so the A10 article is later coverage rather than the original launch. A10 says the integration is available through Wiz's Integrations Marketplace and offers guided onboarding through a TrojAI demo link.
The official company material describes TrojAI as a Canadian technology business with the public offering summarized in this candidate.
Supports: TrojAI by A10 AI security platform · technical tags
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technicaldomainslugs
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technical tags
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technicaldomainslugs
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technicaldomainslugs
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · core features
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · defence applications
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · defence applications
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · core features
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · core features
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technology readiness level
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · summary
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · core features
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · name
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technical tags
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · summary
The official capability material supports the named AI assurance and model security offering and the bounded operational description prepared for reviewer consideration.
Supports: TrojAI by A10 AI security platform · core features
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technicaldomainslugs
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · commercial availability
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technical tags
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · core features
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · technical tags
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · defence applications
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · capability type
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform · defence applications
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform → Autonomous Patrol and Monitoring
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform → Autonomous Patrol and Monitoring
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform → Autonomous Patrol and Monitoring
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
Supports: TrojAI by A10 AI security platform → Autonomous Patrol and Monitoring
Runtime prompt-injection and data-leak controls extend to agent actions, MCP servers and tools, with invocation/span/tool-call telemetry and pre-deployment evaluation. The current offer separates agent runtime protection, coding-agent security, red teaming and model guardrails; vendor efficacy claims require deployment-specific validation.
A10 Networks announced its acquisition of TrojAI Inc. on June 15, 2026 without disclosing transaction terms.
Supports: TrojAI by A10 AI security platform · maturity
A10 announced a completed acquisition of TrojAI on June 15 2026. A10 distinguished build-time AI red teaming from runtime threat protection and said portfolio integration was expected following acquisition, not already complete in that announcement.
TrojAI explicitly announced its Wiz AI-APP integration on March 31, 2026. The article describes Wiz inventory feeding TrojAI testing, findings returning to Wiz and runtime protection through TrojAI Defend; the later September A10 article is not the original integration announcement.
Continue exploring
Follow the strongest connections
Continue through related organizations, reviewed mission and Defence need connections, programme pathways, and explicitly linked intelligence. Similarity results describe shared areas of work, not partnerships or endorsements.
Organizations
Profiles that provide the strongest reviewed or discovery-led continuation.
