Petition for Regulation of Systematic Model Downgrades by Commercial AI Providers

We, the undersigned customers and subscribers of AI services, address this petition to consumer protection authorities and regulatory bodies.
For years, we have witnessed companies such as Anthropic deliberately rolling back or restricting the capabilities of their AI models — often without proper explanation, compensation, or technical necessity. This practice leaves paying customers with products that are no longer the same in quality or performance as what they originally purchased.

These downgrades waste time, energy, and money. They degrade trust and value, turning legitimate technological progress into hidden product deterioration. Despite the commercial nature of these services, consumers currently have no specific legal protections against such regressions.

Artificial intelligence is no longer a “new” field. It is now part of daily digital infrastructure and should therefore meet the same standards of consumer integrity as any other commercial technology. Yet AI-related consumer protection clauses are still largely undeveloped — allowing unregulated downgrades and quietly reduced functionality.

We therefore demand:

The inclusion of explicit “AI Product Integrity” provisions in national and European consumer protection law.
Mandatory transparency regarding functional or performance changes in paid AI services.
Refunds or contractual adjustments when substantial functionality is removed or degraded.
A public hearing and investigation into the regulatory framework for algorithmic downgrades, ensuring accountability for future cases.
This petition stands for the defense of consumers, for technological responsibility, and against covert downgrades that damage trust in AI systems.
Artificial intelligence is no longer experimental — it is a market product. And market products must be regulated.