LEVEL 1AA·PPT·L1
Consumer
Governed access onlyUnderstands AI outputs and uses approved tools within defined parameters. No build capability.
Assessed· Ambient Agent ·
Control
The AI control system for enterprise
AI is already in your organisation. The question is not whether to adopt it. It is whether you control it. Ambient Agent® installs the operating model, governance structure and control system that enterprise AI adoption requires. Deployed in weeks. Designed to be owned and operated by your team.
We do not sell ambient agents. We make AI governance ambient. Ambient Agent is not an AI agent product. We design and install the AI Control System: the access rules, evidence requirements, approval workflow, register and reporting that let an organisation adopt AI safely.
01 / The problem
AI adoption is accelerating without the structures to manage it. Adoption is fragmented across teams, tools and vendors with no central line of sight. The gap is not a knowledge problem. It is a control problem.
No unified view of which AI tools are active, who is using them, or what data they access.
RISKSpend is distributed, duplicated and largely outside financial controls. No consolidated view of AI investment or return.
RISKRegulatory exposure, data leakage and reputational risk grow with every uncontrolled deployment. Audit trails do not exist.
RISKTeams move independently. No shared standards, no common capability model, no governance layer.
RISK02 / The position
Ambient Agent is not a tool. It is an operating model. It sits above your platforms, governing how AI is selected, deployed, measured and scaled.
"You have policy. You do not have enforced control."
The market gap03 / How it deploys
A gated deployment sequence: a two-week diagnostic, a five-stage install of six to twelve weeks, then live governance owned and operated by your team. Every transition passes a control gate. No stage advances on intent.
04 / Agent Passport™
The Agent Passport is the capability-based access model at the centre of the control system. Every person who uses AI in your organisation is passported to the level appropriate to their role and training.
Understands AI outputs and uses approved tools within defined parameters. No build capability.
AssessedConfigures and applies AI within structured workflows. Can adapt approved tools to business context.
AssessedDesigns and deploys AI agents and integrations. Subject to full governance, security review and sign-off.
Assessed05 / Proof
Figures from a deployment inside a UK regulated services organisation. Details are sanitised for commercial use. Independently assessed.
In an independent 8-week maturity assessment.
Baselined with personalised pathways.
One operating model.
Identified and documented from a single intervention.
Design to live, running concurrently.
06 / The regulatory clock
Regulation (EU) 2026/1744, the Digital Omnibus on AI, entered into force on 27 July 2026 and reset some EU AI Act deadlines. It removed none of them. UK regulators are not waiting for statute: accountability expectations apply through existing frameworks today.
Article 4 of the EU AI Act (Regulation (EU) 2024/1689) has applied since 2 February 2025 and requires deployers to take measures to support AI literacy among their staff. It was amended by Regulation (EU) 2026/1744, in force 27 July 2026. National authorities have supervised and enforced it since 2 August 2026. A capability baseline is how you evidence it.
Article 50 of the EU AI Act applies from 2 August 2026. People must know when they are interacting with AI, and AI-generated content must be identifiable. Systems placed on the market before that date have until 2 December 2026 to meet the machine-readable marking duty.
Under Regulation (EU) 2026/1744, in force 27 July 2026, obligations for Annex III stand-alone high-risk AI systems apply from 2 December 2027: risk management, documentation, human oversight, logging.
Obligations for Annex I AI embedded in regulated products apply from 2 August 2028 under the same Regulation. Product-embedded AI joins the same evidence and oversight regime.
A named senior manager is accountable for AI outcomes. AI used in customer decisions must evidence good outcomes and avoid foreseeable harm. Governed through existing frameworks, not a new AI statute.
SI 2026/425, in force 12 May 2026, requires the Information Commissioner to prepare a statutory code of practice on AI and automated decision-making under sections 124A and 124B of the Data Protection Act 2018. The code is in development. Data protection expectations for AI are being codified, not relaxed.
Every regulator asks the same question: can you evidence control at the point of use?
07 / Pricing
No form to fill in before you see a number. The diagnostic is the entry point, and it is credited against the install.
AI Control Diagnostic — £15,000 to £20,000, plus VAT where applicable.
Credited in full against an install within 90 days.
08 / How to begin
A structured two-week assessment that shows you exactly where you are, what is controlled, what is not, and what the system install would look like for your organisation. It creates clarity, not commitment.
What the diagnostic delivers
A complete map of current AI usage, tools and data exposure across your organisation.
A risk posture assessment aligned to your regulatory context and sector requirements.
A capability baseline for your workforce using the Agent Passport framework.
A prioritised value map: where AI can deliver measurable outcome in the next 90 days.
A configured Ambient Agent deployment plan, scoped to your environment and risk appetite.