AGL-US
AI Governance in Law: United States
Assessed against federal guidance and the ABA Model Rules as adopted, with state endorsements layered on the base credential.
Registration and payment are completed on the Solari Institute course platform.
The credential
What it assesses.
- Mark
- AGL-US
- Family
- AGL, AI Governance in Law
- Domain
- AI governance
- Vertical
- Law
- Endorsements
- Status
- Live
The body of knowledge
AGL United States: six domains
The body of knowledge for AGL United States, weighted by share of the examination. The weights are the blueprint: each examination form is built to these proportions, so the emphasis a candidate encounters is the emphasis stated here.
I
Foundations: AI systems, agents, and failure modes
How the systems work in the terms a practitioner needs, and more importantly how they fail. Fabrication, silent error, drift, and the distinct risks introduced when a system is permitted to act rather than only to answer.
10% of the exam
II
Professional responsibility in AI use
The duties that attach the moment a tool touches a matter: competence, confidentiality, communication, candor to the tribunal, supervision of subordinate lawyers and nonlawyer assistance, and fees.
30% of the exam
III
The firm's AI governance program
The NIST AI Risk Management Framework and ISO/IEC 42001 concepts, the governing instruments a firm needs in writing, and the vendor and incident controls that make a policy operate in practice.
20% of the exam
IV
The law governing the lawyer's own AI use
The statutes that reach the lawyer's own AI use, taught as a map and a method for reading new law as it arrives. The firm is treated as the regulated party, a posture most practitioners have not been trained to take about their own practice.
15% of the exam
V
AI work product: verification, ownership, privilege, and records
What has to happen between a model's output and a filing or a client deliverable. Verification that holds up, who owns the result, whether privilege survives the tool, and what the record must show afterward.
15% of the exam
VI
Court-facing and litigation use
Standing orders and disclosure regimes, certification obligations, the sanctions record, and how to run an AI-assisted matter so that the process survives being examined by an opposing party or a court.
10% of the exam
Human rights due diligence runs through the Standard as a connecting thread. Where an AI system affects people who are not the client, the diligence question is a governance question, and the Standard treats it as one.
Competency areas
What a holder must demonstrate.
- The instruments in force in this jurisdiction and how they apply to practice
- Governance structures and decision rights for adopting the domain in an organization
- Supervision and accountability where professional duty cannot be delegated
- Documentation and evidence sufficient to withstand external review
- Escalation and refusal: recognizing when the correct answer is not to proceed
Route to the credential
How a candidate proceeds.
- 01RegisterCandidate registers against the current Standard version for this jurisdiction.
- 02PreparePreparation materials map to the published competency areas.
- 03ExamineAssessment against the Standard, recorded with the version in force.
- 04Register entrySuccessful candidates are recorded with jurisdiction and any endorsements.
- 05MaintainThe maintenance cycle keeps the credential current; lapsed status is recorded.