Attorney General Raúl Torrez and State Representative Linda Serrato announced landmark legislation Wednesday that would create New Mexico’s first comprehensive framework for regulating generative artificial intelligence (GenAI) and synthetic media, commonly known as deepfakes.
Called the Artificial Intelligence Accountability Act (AI2A), the bill aims to protect residents from AI-enabled harm through increased transparency and modern enforcement tools, while preserving free speech and lawful innovation.
Representative Serrato will sponsor the AI2A, which would require providers of generative AI services, large online platforms, and capture device manufacturers to embed latent digital markers in AI-generated images, audio, and video. This would enable victims and law enforcement to track the content’s production and dissemination.
“Generative AI has incredible potential, but it also poses serious risks when used to deceive, defame, or exploit individuals,” Attorney General Torrez said. “The Artificial Intelligence Accountability Act will ensure that accountability and transparency are built into these systems from the start. New Mexicans deserve to know when content is real and when it’s synthetic, and this law would give us the tools to enforce that standard.”
The legislation also empowers the Attorney General to enforce technical provisions of the act and creates civil liability for individuals who knowingly disseminate malicious synthetic content that could harm others. Key provisions include:
- Mandatory Disclosure: AI-generated content must contain latent digital markers identifying it as synthetic.
- Detection Tools: Covered providers must offer free tools to verify the authenticity and origin of digital content.
- Civil Enforcement: The Attorney General would be authorized to investigate violations by AI companies and social media platforms and impose penalties up to $15,000 per violation.
- Enhanced Sentencing: Using generative AI to commit a felony would add an extra year of imprisonment.
Rep. Serrato emphasized the need for state action due to a lack of federal leadership. “For too long, too many New Mexicans have needed protections for both their data and the use of their images,” she said. “I am proud to stand and protect our communities, and to build an AI ecosystem that serves our families instead of harming them.”
The bill has received support from advocacy groups, including Mothers Against Media Addiction (MAMA). Julie Scelfo, the founder and executive director, called the bill a “crucial step to protect children and families” from a growing crisis.
Does this initial draft cover the most important aspects of the press release, or would you like to focus on a particular element for the lead?
This article was drafted with the assistance of Carl, a nm.news AI news research tool, which helped synthesize initial information and research. The final report was thoroughly reviewed, fact-checked, and edited by human editors Noah Gollin & Kevin Hendricks to ensure accuracy and adherence to our editorial standards.

I guess this is a joke? It would have been funny with the italicized postscript, but you left in the paragraph starting with “Does this initial draft cover….?”
Deepfakes are scary, but the scarier thing is that we are allowing ourselves to think less and less by relying on AI to think for us. It’s time to start talking about that part before it’s too late.