United States (California) - Tier A · Elevated · Case: Roadzen
Posted: Sun Oct 11, 2026 4:27 am
Fijishi Jurisdiction Index 2026 · Entry US-CA · Published 12 October 2026
At a glance
1. The rules in United States (California)
The California Department of Insurance's Bulletin 2022-5 (June 30, 2022) names artificial intelligence as a source of bias that insurers and licensees must avoid. Before using any rating or underwriting tool, insurers and licensees must do their own due diligence to ensure full compliance with all applicable laws.
The Department reserves the right to audit and examine insurers' algorithms and models and to take disciplinary action.
Who these rules reach
This band describes how clearly the rules meet the AI use that Roadzen has itself published. It is not a finding that Roadzen has breached any rule.
Why Roadzen: Roadzen's own filing with the US Securities and Exchange Commission says its AI underwriting-decisioning agents entered public beta on May 6, 2026.
The question
3. The same question for other firms in United States (California)
Any insurer or licensee in California that adopts a new AI rating or underwriting tool faces the same point: whether its own compliance due diligence was done for its specific use before it began. This entry gives no band to any firm other than Roadzen.
4. What this entry does not establish
Sources
None share a law or regulator with this entry.
________________________________________
Commercial relationship with any party named here: None · Right of reply: indexreport.protocol@fijishi.com; with the sender's agreement, replies are published in full below the entry, dated · Cite as: Fijishi Jurisdiction Index 2026, Entry US-CA · Corrections: none.
At a glance
| Jurisdiction tier | Tier A - Specified and enforced. A binding instrument that applies to insurers' own use of AI, or an issued regulator document addressed to this insurance market, addresses AI - by name or by an unambiguous description such as "predictive models" or "automated decision system" - and sets out rules or expectations that apply to it; and the regulator has a live mechanism to check or sanction them. |
| Case examined | Roadzen - AI underwriting agents in public beta |
| Band for this case | Elevated |
| Rules reach | Insurers: directly · Reinsurers: not stated · Intermediaries: not stated which licensees; the bulletin is addressed to licensees as well as insurers |
| Last verified | [actual date of the launch-window check] |
The California Department of Insurance's Bulletin 2022-5 (June 30, 2022) names artificial intelligence as a source of bias that insurers and licensees must avoid. Before using any rating or underwriting tool, insurers and licensees must do their own due diligence to ensure full compliance with all applicable laws.
The Department reserves the right to audit and examine insurers' algorithms and models and to take disciplinary action.
Who these rules reach
- Insurers: directly. The bulletin is addressed "TO: All Admitted and Non-Admitted Insurance Companies, Licensees, and Other Interested Parties" (p.1).
- Reinsurers: not stated.
- Intermediaries: not stated which licensees; the bulletin is addressed to licensees as well as insurers.
This band describes how clearly the rules meet the AI use that Roadzen has itself published. It is not a finding that Roadzen has breached any rule.
Why Roadzen: Roadzen's own filing with the US Securities and Exchange Commission says its AI underwriting-decisioning agents entered public beta on May 6, 2026.
The question
In briefRoadzen's own disclosure describes its AI underwriting-decisioning agents as having entered "public beta" the same week the company announced them. California's Bulletin 2022-5, addressed to licensees as well as insurers, says that before using any rating or underwriting tool, insurers and licensees "must conduct their own due diligence to ensure full compliance with all applicable laws". If Roadzen's California-licensed managing general underwriter uses these agents on California policies, has that due diligence been carried out for that use specifically, or does it rest on platform-level testing done before the public beta?
Why this band: The rule in the question binds, names AI and has a fixed date, but whether Roadzen's California-licensed underwriting entity uses the agents is not established.Roadzen's SEC filings call its new AI underwriting agents a "public beta" in the week it announced them. California's insurance regulator requires insurers and licensees alike to do their own legal-compliance due diligence before using any rating or underwriting tool (Bulletin 2022-5, p. 3). No public document says whether Roadzen's California-licensed underwriting entity uses the agents, or whether that due diligence has been done for that use.
3. The same question for other firms in United States (California)
Any insurer or licensee in California that adopts a new AI rating or underwriting tool faces the same point: whether its own compliance due diligence was done for its specific use before it began. This entry gives no band to any firm other than Roadzen.
4. What this entry does not establish
- Whether Roadzen's California-licensed managing general underwriter uses the agents. The question asks this.
- Whether any platform-level testing was done before the public beta. The question offers it as one possible answer.
- The California licence of the managing general underwriter rests on Roadzen's own filings, not on the Department's licence record.
- No date set: the agents leaving public beta.
Sources
- Roadzen Inc., Form 10-K for the fiscal year ended 31 March 2026 (Filing Date 2026-06-29, per the filing index). Company's own. sec.gov
- Roadzen, Form 8-K: "entered public beta on May 6, 2026" (printed). Company's own. sec.gov
- California Department of Insurance, Bulletin 2022-5, June 30, 2022 (printed). Regulator. insurance.ca.gov PDF
None share a law or regulator with this entry.
________________________________________
Commercial relationship with any party named here: None · Right of reply: indexreport.protocol@fijishi.com; with the sender's agreement, replies are published in full below the entry, dated · Cite as: Fijishi Jurisdiction Index 2026, Entry US-CA · Corrections: none.