GLOSSARY
Agentic AI risk, in plain language
The terms your insurer, bank, and board will use about your AI — defined the way one business owner would explain them to another. No jargon, no consulting-deck abstractions.
CORE CONCEPTS
What we mean when we say it
The handful of ideas the rest of the work is built on.
- 01
Agentic AI
The AI tools and assistants that don't just answer questions — they make decisions and take actions on their own: approving, routing, scheduling, placing orders. That autonomy is where the productivity comes from, and where the hidden liability lives.
- 02
Agentic AI risk management
The work of finding where the AI agents in your business create liability, then putting controls, monitoring, and evidence in place so you can scale AI without scaling risk. It's a continuous programme with a named owner — not a one-time project.
- 03
AI risk register
A live record of every AI agent and tool you run — what each one can access, how much it can act on its own, and the controls on it. It's the source of truth your insurer, board, and bank can read when they ask, and the evidence behind staying insurable and fundable.
- 04
Insurable and fundable
The position we keep you in. Insurable: your commercial insurer covers your AI exposure without exclusions or surcharges. Fundable: your bank, your covenants, and the grants and tax credits you qualify for get a clear, documented read of your AI risk. We improve your odds on insurer and lender readiness; for grant and tax-credit filings, we map what's available and refer you when you don't have someone in-house. The decision stays with the bank, lender, or insurer.
- 05
AI investment strategy
The first piece of work we do: where AI actually pays off in your operation, what to avoid or defer, and a prioritised roadmap — with the grants, tax credits, and financing to fund it. The money question comes before the build.
- 06
AI risk diagnostic
A first read of the AI you already run and the exposure it carries — the starting point your risk register and investment roadmap are built on.
- 07
Managed AI risk
Your outsourced AI risk team on a monthly subscription: we keep your register current, watch for threats that hit your specific agents, check every new agent before it goes live, and keep your funding and insurance readiness up to date — the capability without the in-house headcount.
- 08
Continuous AI risk monitoring
An ongoing watch for new threats, regulatory changes, and shifts in best practice — checked against your register so you hear about what actually affects the agents you run, and what to do about it.
PROOF, NOT PITCH
The references your banks and insurers actually trust
We don't sell frameworks and regulations — we cite them as proof the method holds up when your insurer, lender, or board asks. Below: the standards worth knowing, the money on the table, and a few of the risks worth naming plainly.
FRAMEWORKS, FUNDING & RISKS
Terms worth knowing
- 01
AI insurability
Whether your commercial insurer will cover your AI exposure, and on what terms. Insurers are starting to add AI exclusions to commercial policies; documented controls and evidence from your own systems let an underwriter price the real risk instead of excluding it.
- 02
NIST AI Risk Management Framework
A widely recognised, voluntary framework from the U.S. National Institute of Standards and Technology for identifying and managing AI risk. We align our work to it as proof the method is sound — not a badge we sell.
- 03
ISO/IEC 42001
The international management-system standard for artificial intelligence — the recognised reference for governing AI responsibly. We build to it where it matters to your insurers and lenders.
- 04
Loi 25 (Quebec)
Quebec's modernized private-sector privacy law. When AI tools touch customers' personal information, Loi 25 sets obligations — including mandatory notification when a confidentiality incident happens — that your AI operations have to account for.
- 05
EU AI Act
The European Union's risk-based law for AI. Even outside Europe it shapes what insurers, partners, and regulators expect — a reference point we track so your operations don't get caught flat-footed.
- 06
SR&ED and AI tax credits
Tax credits — like the federal SR&ED programme and Québec's R&D credits — that AI and technology work often qualifies for, but many businesses never claim because they don't know they're eligible or can't document it. We find what applies and document the work in your register so it holds up when you or your advisor claims it.
- 07
AI-adoption grants
Government and local-agency programs that help fund AI adoption — real money in Québec and Canada. We identify the ones you qualify for and build them into your funding plan — filing stays with you or a referred specialist.
- 08
Hallucination
When an AI states something false with full confidence. Harmless in a rough draft; costly when it feeds a real decision — an email to a customer, a filing, a price. One of the concrete risks a register and tested controls are there to catch.
Not sure which of these apply to your business?
Book a short, confidential call. Tell us what AI you're running or planning, and we'll translate it into where you stand — what's exposed, and what it takes to stay insurable and fundable. We'll tell you if it's not a fit.