Data Under Review
This bill's information is currently being re-verified for accuracy. Some details may have changed since our last confirmed update. For the most current information, please check your state legislature's official website.
Disclaimer: This page provides general informational summaries only and does not constitute legal advice. AI-generated content may contain errors. Always consult a qualified attorney for guidance specific to your situation.
Read full disclaimer →
AI Summary
The bill mandates developers of AI models to implement safety protocols and report on critical risks.
Business Impact
If you develop AI models in Michigan, you must implement safety protocols and publish reports by January 1, 2026, or face fines up to $1 million.
What do these statuses mean? ▼
Introduced
— Filed in the legislature; not yet heard in committee
In Committee
— Assigned to and being reviewed by a legislative committee
Passed
— Approved by one or both chambers; awaiting further action
Signed / Enacted
— Signed into law by the governor; may or may not be in effect yet
Dead / Vetoed
— Vetoed, failed to pass, or session expired without action
Unknown
— Status data not yet available or awaiting classification
Affected Industries
Topics
What This Means
House Bill 4668 requires developers to manage critical risks associated with AI models. Specific timelines for safety protocols and transparency are not confirmed.
Key Provisions
- Developers must implement and publish safety and security protocols.
- Developers must produce transparency reports detailing risk assessments.
- Civil fines for non-compliance are not confirmed without the full bill text.
- The duration for record retention is not confirmed without the full bill text.
- Protection from retaliation for employees reporting critical risks is not confirmed without the full bill text.
- Developers must adhere to the specified timeline for publishing modifications to safety protocols is not confirmed without the full bill text.
Latest Legislative Action
Referred To Committee On Communications And Technology
Bill Sponsors (showing 30 of 61)
| Name | Role |
|---|---|
| Jay DeBoyer R | Primary |
| Sarah Lightner R | Primary |
| Tommy Brann | Primary |
| 39 | Cosponsor |
| Adam Zemke | Cosponsor |
| Alicia St. Germaine R | Cosponsor |
| Andrew Beeler | Cosponsor |
| Angela Rigas R | Cosponsor |
| Beth Griffin | Cosponsor |
| Brandt Iden | Cosponsor |
| Brenda Carter D | Cosponsor |
| Brian BeGole R | Cosponsor |
| Cam Cavitt R | Cosponsor |
| Cara Clemente | Cosponsor |
| Curtis VanderWall | Cosponsor |
| Darrin Camilleri D | Cosponsor |
| Dave Prestin R | Cosponsor |
| David Martin R | Cosponsor |
| David Maturen | Cosponsor |
| Denise Mentzer D | Cosponsor |
| Donavan McKinney D | Cosponsor |
| Donni Steele R | Cosponsor |
| Dylan Wegela D | Cosponsor |
| Eric Leutheuser | Cosponsor |
| Gregory Alexander | Cosponsor |
| Holly Hughes | Cosponsor |
| James DeSana | Cosponsor |
| John Roth | Cosponsor |
| Jon Hoadley | Cosponsor |
| Joseph Aragona | Cosponsor |
Compliance Checklist
Implement and publish a safety and security protocol
Who: Large developers of AI models
Deadline: By January 1, 2026
Penalty: Up to $1 million per violation
Who: Large developers of AI models
Deadline: By January 1, 2026
Penalty: Up to $1 million per violation
Produce and publish transparency reports every 90 days
Who: Large developers of AI models
Deadline: Every 90 days starting January 1, 2026
Penalty: Up to $1 million per violation
Who: Large developers of AI models
Deadline: Every 90 days starting January 1, 2026
Penalty: Up to $1 million per violation
Related & Companion Bills
Full Legal Analysis
🔒
Full Analysis for Subscribers
Get the complete legal breakdown, compliance checklist, enforcement timeline, and expert commentary on HB4668.
Unlock Full Analysis →Official Source
Official Bill Text
View HB4668 on official legislature website →
Related Topics
Affected Industries
More Michigan AI Legislation
HB 4537
Human services: medical services; use of artificial intelligence tool…
Medium
HB 4668
Trade: business regulation; requirements and safety standards for dev…
Medium
HB 4536
Insurance: health insurers; use of artificial intelligence tools to m…
Medium
HB5579
Labor: fair employment practices; use of electronic monitoring or aut…
High
HB5141
Campaign finance: advertising; using artificial intelligence in certa…
Medium
HB4667
Crimes: other; use of artificial intelligence to commit certain crime…
High
View All MI Laws →
More Michigan AI Laws
Browse all AI bills and regulations tracked for Michigan.