AI governance and evidence
Trace extraction proposals, human decisions, access labels, and retrieval sources through a governed AI workflow.
Start here
- Audit Trails for AI Document Extraction
Follow a source document, an AI proposal, and a corrected business value.
- Should AI Labels Control Document Access?
Separate classification suggestions from access authority.
- Protect RAG Against Poisoned Documents
Protect retrieval and tool actions from untrusted document content.
More on ai governance and evidence
- Audit Trails for AI Document Extraction
Connect each accepted AI-extracted value to its source document, model proposal, reviewer decision, and later corrections.
- Australia's Automated-Decision Privacy Policy Changes
Map document inputs, automated decisions, human review, and privacy-policy disclosures for Australia's automated-decision requirements.
- Bring AI into Document Management with Controlled Processing
Intelligent document processing uses software to classify documents and extract information for a business workflow.
- Canada's OSFI AI Guidance: Govern Document Agents
Apply OSFI AI guidance to document agents in Canadian financial institutions, with scoped actions, reviewed decisions, and fallback procedures.
- Canada's OSFI E-23: Model Change Evidence
Prepare for Canada's OSFI E-23: link model validation, change approval, release evidence, and monitoring for traditional and AI models.
- Checking Invoice Rows After OCR
Check invoice table rows after OCR: match quantities, units and amounts, retain source context and review exceptions before using accepted data.
- Colorado ADMT: Prepare Decision Evidence for 2027
Prepare Colorado ADMT evidence linking document inputs, system versions, human review, notices, and appeals as rulemaking develops.
- Document Deletion Across Search, Previews, and AI
Carry approved document deletion through search indexes, previews, and AI retrieval, with explicit scope and reconciliation.
- EU AI Act: Evidence for Document AI Workflows
Connect document AI inputs, model configurations, human decisions, monitoring, and retained evidence to the workflow's EU AI Act role.
- FormKiQ KnowledgeBase: AI Retrieval in Your AWS Account
See how FormKiQ KnowledgeBase uses Amazon Bedrock for natural-language retrieval across governed document collections in the customer’s AWS account.
- Improve Document Workflows with Measured Automation
Automate a document workflow when you can define the required outcome, recognise exceptions, and measure the result.
- Protect RAG Against Poisoned Documents
Protect document retrieval and AI actions against poisoned sources, prompt injection, permission leakage, and stale content.
- Should AI Labels Control Document Access?
Separate AI sensitivity proposals from access decisions, with trusted labels, reviewed changes, and safe handling of uncertain classifications.
- Trustworthy AI Workflows: Validation and Human Review
How deterministic validation, cross-checks, human review, prompt and model change control, and audit history make AI document workflows more trustworthy.
- Use AI to Assist Process Design
AI can help turn a written business process into a candidate workflow, schema, or field mapping.
- What We Learned About AI for Document Management
Our early document-AI experiments showed that useful demonstrations and dependable workflows are different milestones.
- When a Knowledge Base Cannot Answer
Handle unanswered document questions with clarification, an owned review queue and verified fixes using FormKiQ KnowledgeBase.
- Which Policy Should Your AI Assistant Use?
Connect FormKiQ policy approvals to AI answers. Separate current, upcoming and historical guidance, inspect source revisions and verify each policy change.