Question · DoW RAI S&IP

What does the CDAO RAI Toolkit / GenAI Guardrails require for fielded models?

AboutDoW RAI S&IP

Legal forceInternal departmental strategy; directs DoD components, no direct duty on contractors

Applies fromJune 2022

Checked21 Oct 2026

01 / The answer

The short answer

Neither imposes binding requirements or numeric thresholds on fielded models; both are internal DoD guidance, binding on contractors only where a contract says so. The CDAO Responsible AI Toolkit, released in 2023 and now titled the AI Assurance (AIA) Toolkit, guides teams through identifying, tracking and documenting AI risks across the life cycle against the DoD (now styled DoW) AI Ethical Principles. CDAO's July 2024 memo, Guidelines and Guardrails to Inform Governance of Generative AI, addresses issues such as data quality, cyber and adversary threats, and traceability. The January 2026 AI strategy memo directs CDAO to realign this guidance, so both may change.
DoW RAI S&IP · checked 21 Oct 2026

02 / The rule

What DoW RAI S&IP requires

The June 2022 Pathway turns DoD's five AI Ethical Principles into 64 lines of effort organized under six tenets: governance, warfighter trust, product and acquisition lifecycle, requirements validation, ecosystem, and workforce. It is the basis for the CDAO's Responsible AI Toolkit and for how DoD programs are expected to document trust in fielded AI.

01Apply the DoD AI Ethical Principles (responsible, equitable, traceable, reliable, governable) across the AI lifecycle.
02Designate Responsible AI leads in each DoD component and resource the lines of effort assigned to that component.
03CDAO provides department-wide Responsible AI guidance, tools and coordination, and tracks implementation progress.
04Build warfighter trust through test and evaluation, assessment and monitoring of AI capabilities.

03 / Records

Records to keep

01Completed RAI Toolkit (now AI Assurance Toolkit) assessments for the AI project
02Test and evaluation records showing reliability of the fielded model
03Traceability documentation: data sources, design decisions, model versions
04Component Responsible AI lead designation and line-of-effort status reports

Optional evidence · where Planisphere fits

Keep a record anyone can check

When your team meets a duty, Planisphere can seal the event it reports as a receipt.

01 · Your teamRecord the event

A disclosure shown, a review signed, a model change approved: the event your system reports.

02 · PlanisphereSeal it

The event becomes a sealed receipt, dated and attributed to whoever is signed in.

03 · AnyoneCheck it

Paste the receipt into Verify. No account or key.

A record supports your compliance evidence process. It is not a legal finding or a certification. Systems that call Planisphere's gate API can also hold or refuse an action before it runs.

SOURCES