Louisiana AI Legislation in 2026

The Louisiana Picture

Louisiana has not enacted a comprehensive artificial intelligence statute, and as of June 2026 it has no plans to. What the state has built instead is a cluster of narrow, harm-specific criminal and procedural laws aimed at the most acute abuses of synthetic media. I see three pillars. First, sexual deepfakes and child exploitation: Act 457 of 2024 criminalized the creation and distribution of sexual deepfakes with severe, parole-ineligible penalties when a minor is depicted, and the 2026 Senate Bill 42 went further by rewriting the state's child sexual abuse material definitions so that AI-generated and virtually indistinguishable images count. Second, evidentiary integrity: Act 250 of 2025 made Louisiana one of the first states to put rules around AI-generated evidence in court. Third, a study-and-recommend posture through HCR 66 of 2024.

The distinctive feature of Louisiana's approach is its restraint on the regulatory side and its aggression on the criminal side. The state has repeatedly declined to regulate AI as an industry. The governor vetoed the 2024 political deepfake and watermarking bills as threats to political speech, and in the 2026 session roughly a third of nearly twenty proposed AI guardrail bills were withdrawn after a federal executive order signaling that states regulating AI could lose federal funding. The measures that survive are the ones framed as protecting children, protecting courts, and punishing clearly malicious conduct, not the ones that impose disclosure, bias-audit, or licensing duties on developers and deployers. This is a prosecution-first, regulation-light model.

For builders and deployers, the practical implications are concrete even without a broad AI statute. If you generate, host, or distribute synthetic imagery, you face real criminal exposure in Louisiana the moment that imagery is sexual and especially when it could depict a minor, and the SB 42 changes mean that the image was synthetic or no real child exists is not a defense. If your platform handles user-generated media, the realistic course is proactive detection and takedown of sexual deepfakes and CSAM. If your product touches litigation, e-discovery, or forensics, you should be prepared to support authenticity verification and provenance, because the duty now sits on the attorneys who rely on your output.

My guidance is to treat Louisiana as a high-enforcement, low-prescription environment. Do not wait for a comprehensive compliance regime that is unlikely to arrive soon, and do not assume the absence of a broad AI law means low risk. The risk here is criminal and reputational rather than administrative. Watch the Joint Legislative Committee on Technology and Cybersecurity and its annual reports for the direction of future bills, watch the final enactment status of SB 42 and the related training-data bill, and build provenance, consent, and abuse-prevention controls now so that you are aligned with the harms Louisiana has chosen to punish.