Circular 230 sets the standards of practice for those who practice before the IRS — CPAs, enrolled agents, and attorneys. It demands competence, due diligence as to accuracy, and reasonable reliance on information. An AI that fabricates a citation or a figure is a Circular 230 problem. Cortex grounds every figure in a cited source, holds the return for a practitioner's review, and seals the work product in a ledger that proves the diligence was done.
Built for Treasury Circular 230 · §10.22 diligence · §10.35 competence
Circular 230 (31 CFR Part 10) governs practice before the Internal Revenue Service. Key duties include §10.22 (diligence as to accuracy of returns and documents), §10.35 (competence — the knowledge, skill, and preparation appropriate to the matter), and §10.34 (standards for advising on positions). A practitioner may generally rely in good faith on information from a client, but must make reasonable inquiries when it appears incorrect or incomplete. For AI in tax workflows, the failure modes are hallucinated authorities and ungrounded numbers. Cortex's answer is grounded, cited generation: every figure on the return links to a source document, page, and box; the return is held for the practitioner's review before filing; and the whole chain — figure, source, review, approval — is sealed and provable.
applies to ▸ United States · tax practitioners practicing before the IRS
Circular 230's diligence and competence standards reduce to one operational test: can the practitioner prove the work was grounded, reviewed, and accurate? Cortex makes each step a recorded control.
Every obligation reduces to a fail-closed runtime gate whose verdict lands in a tamper-evident ledger you can hand an examiner. This is the table a Compliance Pack exports for this framework.
| Obligation | Enforced Cortex control | Ledger evidence |
|---|---|---|
| §10.22 — diligence as to accuracy | Datapoint provenance + citations | page · bbox · asReported |
| Reasonable reliance / inquiry on data | value_hash + grounded generation | ungrounded figure flagged |
| Human review before the return is filed | Approval gate + Oversight | 409 hold · reviewer receipt |
| §10.35 — competence & quality of work | Gold evals + reliability score | 97% accuracy · 95% coverage |
| §10.34 — standards for advising positions | Policy-as-Code thresholds | require_approval ≥ threshold |
| Work product is complete & tamper-evident | Signed receipts, sealed chain | hashOk:false flags edits |
If the IRS questions a position, the practitioner must show the work was grounded and diligent. A Compliance Pack maps the Circular 230 duties to their Cortex controls and exports the sealed work product — figure, cited source, the practitioner's recorded review, and the approval — verifiable offline.
These are not slideware promises — they are the literal codes and receipts the runtime returns when you challenge it against this framework's controls.
Pick IRS Circular 230. Cortex lines each obligation up against the runtime gate that enforces it — no spreadsheet archaeology.
Every agent run passes the same fail-closed gates. A denied control returns a real code (402 / 403 / 409) — never a silent pass.
Export a Compliance Pack: the mapping table plus the sealed ledger records that show each control fired, verifiable offline.
Circular 230 governs practitioners, not software, and there is no certification for a vendor. Cortex is built for and aligned with its diligence and competence standards by grounding, reviewing, and sealing tax work product; professional responsibility remains with the practitioner.
Each obligation above is enforced by a real capability in the runtime. Explore the ones that do the work for this framework.
Datapoint provenance pins every figure to its source filing, page, and box — the diligence proof Circular 230 expects.
The approval gate holds the return for the practitioner's review before filing.
A signed, SLA-backed tax pack with a gold eval suite and grounded, cited generation.
The same enforced controls and one-click evidence export cover the other frameworks your auditors cite.
Aligned with FINRA rules: books-and-records, supervision, and communications review — with a tamper-evident, retained audit trail..
Aligned with SOC 2 Trust Services Criteria: access controls, change management, audit logging, and tamper-evident evidence..
Aligned with the EU AI Act: human oversight, logging and traceability, risk management, and audit-ready evidence packs..
The full compliance hub — every framework Cortex maps to, with the shared control-to-evidence model.
The sealed ledger, signed receipts, and lineage graph map to the obligations across every regime you report against — aligned with, never claiming a certification you don't hold.
See how Cortex maps IRS Circular 230 to enforced controls and exports auditor-grade evidence on demand.