Mexico Drafts AI Law With Prison Terms for Deepfakes

Mexico is stepping up where many countries have hesitated, drafting sweeping artificial intelligence legislation that balances innovation with serious protections against abuse.

The Mexican Senate's AI committee unveiled a comprehensive bill after 16 months of work involving 72 specialists from universities, businesses, human rights groups, and government agencies. All political parties in Congress participated in shaping the legislation.

The proposed law tackles some of AI's most dangerous applications head-on. It would ban creating or sharing sexually explicit deepfakes without consent, with absolute prohibition for content involving minors regardless of consent. Prison sentences would apply to violations.

The bill also prohibits using AI to automate harassment campaigns, create deepfakes designed to humiliate or extort people (especially targeting women), and develop autonomous weapons systems that operate without human oversight. Election manipulation using AI would face similar restrictions.

To prevent confusion about what's real online, the legislation requires transparency labeling when AI has been used to create or modify content. This helps readers and viewers know when they're seeing something computer-generated versus authentic.

The Bright Side

Mexico isn't just focusing on what could go wrong. The bill proposes a national AI strategy, a development fund to support innovation, and a federal certification system to help companies build AI responsibly.

Senator Rolando Zapata, who chairs the AI committee, emphasized that the law protects democratic debate rather than restricting it. "This project in no way limits or restricts freedom of expression or political debate," he explained on social media.

Some cybersecurity experts worry that vague terms like "cognitive manipulation" could be interpreted too broadly by authorities, potentially affecting free speech and innovation. The Senate will now debate these concerns as the bill moves through the legislative process.

Mexico's comprehensive approach contrasts sharply with the United States, which has pursued AI regulation piece by piece without unified federal legislation. If passed, Mexico's law could become a model for other nations wrestling with how to harness AI's benefits while protecting citizens from its risks.

The bill represents a growing recognition worldwide that artificial intelligence's rapid expansion requires thoughtful guardrails, not just optimism about its potential.