The factory: one infographic per company
For every company we want to meet at Web Summit: what it does, the risks it faces from AI agents, and how RiskMandate helps, each in normal terms and in lay terms side by side. To help the team understand what we are doing and why each conversation matters.
RiskMandate in one picture
- Reachwhat it can actually access or affectEverything an AI agent can actually access or affect: systems, data, actions, connectors.
- Mandatewhat the business authorisedWhat the business has authorised it to do: the task, the scope, the purpose.
- Gapthe difference: the riskWhat the agent could do but was never asked to. This is where the risk lives.
- Barrierswhat really stands in the wayThe real controls that stop unwanted behaviour, each with evidence: technical, rule or review.
- Provenancewho owns it, who reviewed it, which versionWho owns the agent, who reviewed it, and which version is running.
In normal terms
RiskMandate gives every AI agent an Agent Behaviour Policy: a factual record of its Reach, its Mandate, the Gap between them, the Barriers that hold, and its Provenance, ending in a Licence to Operate. Operators, security, leadership and insurers each read the same record from their own side. The line: know their reach, define their mandate, control the gap.
In lay terms
An AI agent is like a new employee with a master key. RiskMandate writes down which doors the key opens, which rooms the employee is actually meant to be in, and what stops them wandering into the others. Then everyone, from the boss to the insurer, can see it on one page.
The companies (31)
Cisco (Outshift by Cisco)ai-agents · United StatesCisco's incubator and developer platform for agentic AI: building the plumbing that agents will run on.Hexamindai-agents · FranceAn open-source gatekeeper for every tool an AI agent calls: the closest neighbour to what we do.Primitiveai-agents · United StatesEmail built for AI agents: each agent gets its own address and can send and receive like a person.Mastradevtools · United StatesA leading open-source TypeScript framework for building production AI agents.Qasedevtools · United StatesTest management for 2,000+ engineering teams, now checking every change a coding agent makes against what was asked.Replitdevtools · United StatesAnyone can build an app by describing it; the agent writes, runs and ships the code.KPMGgrc · PortugalAudit, tax and advisory: the people boards trust to say whether risk is under control.Startup Portugalportugal-startup · PortugalPortugal's startup agency: the door to the Portuguese startups at Web Summit.1Passwordsecurity · CanadaPasswords and secrets for 150,000 businesses, now including the AI agents that log in on their behalf.Cloudflaresecurity · United StatesA connectivity cloud and security platform, now saying 'build for the agent era'.Delineasecurity · United StatesIdentity security that governs what people, machines and AI agents do once they are in.Polygraf AIsecurity · United StatesA behavioural control plane that watches and enforces what humans and AI agents do, live.Descopesecurity · United StatesOne identity platform for people and AI agents: logins for humans, scoped access for agents and MCP servers.Airiagrc · United StatesOne governance layer for AI: agents, models, tools, data, runtime policy and audit evidence.Emergenceai-agents · United StatesAgentic infrastructure for enterprises: verified, governed agents. Speaking on 'When AI agents go rogue'.Feedzaisecurity · PortugalPortugal's AI-native fraud prevention scale-up, speaking on 'Who let the agent in?'Linux Foundation (Agentic AI Foundation)ai-agents · United StatesA neutral home for open standards in agentic AI.Consciumai-agents · United KingdomAn AI safety lab that tests and verifies agents in simulation before they are deployed.Scrut Automationgrc · United StatesCompliance automation for SOC 2, ISO 27001 and 60+ frameworks, run with AI agents.Cnextai-agents · BelgiumBuilds secure, compliant agents for European businesses on its own Kairos platform.Risiq GmbHgrc · GermanyRisk and regulatory advice for banks, with a platform for ESG and risk reporting.Melious AI GmbHplatform · GermanySovereign European AI: frontier open-weight models served from Europe, outside the US Cloud Act.ChatPlaceportugal-startup · PortugalA Portuguese AI manager that answers creators' messages and comments, 24 hours a day.Firefox Enterprise by Mozillasecurity · NetherlandsA customer-controlled enterprise browser for European organisations that want control of their data.SpaceComputersecurity · SingaporeConfidential computing, keys and attestation for cloud and AI workloads, up to satellites.Trent AIsecurity · United KingdomAn AI security engineer that finds the real risk in code, cloud and AI agents, and verifies each fix.TrustScalegrc · United StatesMeasures AI against a customer's own criteria: training, evaluation and assurance.AICEP - Portugal Trade & Investportugal-startup · PortugalPortugal's trade and investment agency: in about 50 countries, bringing investment in and taking Portuguese companies out.GetVocal AIai-agents · FranceAI agents that run large companies' customer contact centres, sold on control and trust.PolyAIai-agents · United KingdomVoice AI agents that answer customer phone calls for large companies, from start to finish.Legoraai-agents · SwedenAI and agents for lawyers: reviewing, researching and drafting legal work.