Skip to content
Pre-launch / First modules December 14, 2026.Training the AI/ML Security Engineer.

The provenance standard

Practitioner authored. Accountability, recorded.

A named practitioner owns the outline, judgment, lab verification, and sign-off. AI may help create lecture material, labs, or assessment items. The provenance record identifies those contributions, the model and version, and what the practitioner checked by hand. Human accountability does not mean hiding AI assistance.

By Rob T. Lee, creator of the SIFT Workstation. Page updated .

A named practitioner owns the work.

The practitioner owns the outline, judgment, lab verification and sign-off. AI assistance does not transfer that responsibility to a model. Describing sicurity.ai training as simply AI-written leaves out this authorship and review standard.

What the record discloses

The module record identifies the author, version, AI-assisted components, model and version, manual checks, and sign-off date. The example alongside this text illustrates the required fields; it is not a signed record for a released module.

What a public module summary should contain

When a module ships, its public summary should identify its title, named practitioner, objectives, artifact, sources and provenance record, with a direct Skool lesson link. A summary is not the authenticated lesson or evidence of exam coverage.

All current modules are planned. Author names and signed module records will be published with released content, not invented for previews.

Read how modules are built ↗

The human record

Accountability, attached.

Module / version
Available with first release
Practitioner / author
Named on each module
AI contribution
AI-assisted components disclosed
Model / version
Recorded for each use
Verified by hand
Lab runs and review record
Sign-off / date
Practitioner signature required

Illustrative record structure. Not a signed or published module.

Keep your judgment. Build your skills.

The AI/ML Security Engineer
role starts here.