Business Questions
Find the answer that matters to your business decision. Explore common questions about reporting, accountability, AI, information readiness, adoption, and DACS conformance. Each answer points you to a focused DACS page and, where useful, a supporting document or example.
Choose your business concern
Select a topic to jump to its questions, or scroll to browse all six categories.
Trust and reliability — Can we rely on the information?
Accountability and risk — Who is responsible, and what happens when something fails?
AI and automation — When can AI results be used or acted upon?
Governance, meaning and evidence — What must we know and prove?
Getting started — How can we begin without a large program?
Assurance and conformance — What can we credibly claim?
These are orientation answers. For formal control requirements and evidence obligations, use the DACS Specification.
Trust and reliability
Can we rely on the information?
Can we trust the numbers in a board report?
A number can be technically correct yet unsuitable for the decision being made. The definition, population, period, permitted use, and supporting evidence need to match the report’s stated purpose. DACS helps make that reliance question explicit.
Start here: Where Defensible Analytics Matters Go deeper: DACS First Look
What if two departments report different versions of the same metric?
First identify whether the measures use the same meaning, grain, population, and time boundary. If they do not, apparent disagreement may be a definition conflict rather than a calculation error. A governed semantic contract makes those distinctions visible.
Start here: Semantic Contracts Go deeper: System of Information Management (SIM)
Is passing data-quality checks enough to trust an analytical result?
Not by itself. Data quality is useful evidence, but it does not independently establish that a result is appropriate or authorized for a particular business decision. The decision scope and other applicable requirements still need evaluation.
Start here: What is Defensible Analytics? Go deeper: Existing Data and AI Controls
What happens if evidence changes after a decision is approved?
The basis for reliance must remain reviewable. When material evidence, context, or conditions change, the affected use may need reassessment. An earlier readiness decision should not be treated as permanent permission.
Start here: DACS Operating Model Go deeper: Accountability Gates
Accountability and risk
Who is responsible, and what happens when something fails?
Who is accountable for using an analytical result?
Responsibility should be established for the particular decision or use, rather than assumed from ownership of a dataset or platform. DACS distinguishes evidence production from authority to accept or restrict reliance on that evidence.
Start here: DACS Framework Go deeper: Accountability Gates
What happens when a required control fails?
The failure must remain visible. A Gate decision can restrict or stop the affected use, or allow constrained use where an authorized exception is supported; an exception does not change a failed control into a pass.
Start here: Accountability Gates Go deeper: Filled Accountability Gate Example
What should the board ask management to demonstrate?
For a material decision, the board can ask who owns the use, which requirements apply, what the evidence supports, which issues remain open, and what action is permitted. The purpose is inspectable oversight, not a blanket claim that every system is safe.
Start here: Where Defensible Analytics Matters Go deeper: DACS Framework
Can we reconstruct why a decision was allowed?
The goal is to preserve a scope-bound record of applicable controls, evidence references, results, readiness disposition, exceptions, and review conditions. The record may be kept in existing systems if it remains inspectable.
Start here: DACS Operating Model Go deeper: DACS Specification
AI and automation
When can AI results be used or acted upon?
How do we decide whether an AI output may influence a business decision?
Start with the specific use and its consequences. Determine what AI-related context and controls are material, what evidence exists, who has authority, and whether the resulting readiness decision permits, restricts, or stops the use.
Start here: DACS Profiles Go deeper: Accountability Gates
Can an AI agent act automatically without a human approving every step?
The relevant issue is the authority granted for the particular action, together with constraints, applicable evidence, and escalation or stopping conditions. DACS provides a way to evaluate the controlled use; it does not imply that every agent action is permitted.
Start here: DACS Operating Model Go deeper: DACS Specification
Do our existing AI guardrails and monitoring tools already satisfy DACS?
They may contribute important evidence, but an operating tool is not automatically the scope-bound readiness decision. Map the controls you already have to the applicable requirements, and identify any gaps without assuming a replacement platform is necessary.
Start here: Existing Data and AI Controls Go deeper: DACS Profiles
Governance, meaning and evidence
What must we know and prove?
Who decides what a business term means?
The governing definition, accountable authority, identity, grain, scope, and permitted use should be explicit wherever they affect a material decision. Semantic Contracts provide a way to bind those elements rather than rely on informal agreement.
Start here: Semantic Contracts Go deeper: SIM and SIM-ZONEs
Why do we need SIM if we already have data governance?
SIM describes information management responsibilities and readiness contexts. It helps frame how information is organized for use, while DACS evaluates applicable controls for a stated decision path. The two concepts contribute different responsibilities.
Start here: System of Information Management (SIM) Go deeper: SIM and SIM-ZONEs
Do we have to apply every DACS control to every decision?
No. The Core Decision-Scope Controls apply to every DACS evaluation, while additional profiles are selected when their subject matter is material to the declared use. The applicability rationale should remain explicit.
Start here: DACS Profiles Go deeper: DACS Specification
Getting started
How can we begin without a large program?
Can we start with one important business decision?
Yes. A bounded decision scope is the intended starting point for exploring whether evidence, accountable ownership, and a readiness decision can be made explicit. The First Look helps discover what already works and what remains unresolved.
Start here: DACS First Look Go deeper: DACS Adoption Roadmap
What does a completed First Look actually look like?
An example helps make the deliverable tangible: declared scope, evidence reviewed, findings, unresolved questions, proposed restrictions, and follow-up actions. The illustrative example is not a completed formal conformance assessment.
Start here: Completed First Look Example Go deeper: DACS First Look
Who should participate, and what should each person read?
Participation depends on the decision scope and the skills required; job titles alone do not determine accountability. The role-based guide provides shorter reading paths for executives, governance leaders, risk teams, architects, and implementers.
Start here: DACS by Role Go deeper: DACS Framework
How do we move from a First Look to repeatable adoption?
Use the initial assessment to determine what needs remediation, which responsibilities must be assigned, and what can be repeated. Expand only where control evidence and accountable decisions support broader use.
Start here: DACS Adoption Roadmap Go deeper: DACS Operating Model
Assurance and conformance
What can we credibly claim?
Does DACS certify our company, product, model, or platform?
No. DACS conformance concerns an explicitly stated decision scope and applicable evidence-backed requirements; it is not blanket certification of an enterprise, person, vendor, platform, or model. Claims should preserve that boundary.
Start here: DACS Operating Model Go deeper: DACS Specification
Where are the formal requirements defined?
The public site explains concepts and use cases; the DACS Specification is the normative reference for requirements, applicability, evidence, and conformance boundaries. Use it when the precise control language matters.
Start here: DACS Specification Go deeper: DACS Library
Can we use evidence from tools we already operate?
Potentially. Existing catalogs, lineage, quality, governance, observability, and AI controls can contribute evidence. What matters is whether that evidence meets the applicable requirements for the declared use and can support an inspectable readiness disposition.
Start here: Existing Data and AI Controls Go deeper: DACS Specification
Prefer to navigate by role?
If you know your responsibility but not where to begin, choose a role-based reading path. Or explore how the major parts of DACS fit together.