ARTICLE I
CONTROL INPUT AND OUTPUT

I.
All organizations, whether private or state, engaged in designing, engineering, and/or distributing AI products, shall be held unequivocally accountable for the information generated by AI systems.
II.
These organizations must establish specialized departments amalgamating human oversight and automated technologies grounded in machine learning to guarantee the responsible utilization of AI.
III.
An external, impartial global agency shall meticulously oversee and ensure strict adherence to proper AI usage, conferring AI-Safe-Use approval badges exclusively upon organizations that diligently comply with AI standards.

ARTICLE II
TRANSPARENT TRACING OF SOURCES

I.
Complete transparency in acknowledging the entities responsible for AI processes is imperative. Therefore, all AI-processed information must be transparently traceable to its origins, specifically attributing it to the entities conducting the information processing using AI.
II.
Users shall enjoy unrestricted access to all original input data employed by AI systems.
III.
Violations of source tracing obligations will be met with resolute legal enforcement.

ARTICLE III
REGULATION OF DEEPFAKES

I.
Mandatory watermarks or detectable patterns are recommended for all deepfake or artificially created content.
II.
We advocate increased investment in deepfake detection technologies.
III.
Unethical deepfake actions, including defamation and identity theft, must be unequivocally prosecutable offenses.
IV.
AI systems shall rigorously maintain accessible interaction histories, with AI software manufacturers being legally accountable for verifying the origin of disseminated information.

ARTICLE IV
PREVENTION OF MONOPOLIES AND DUOPOLIES

In the pursuit of equitable AI development and access, signatory parties solemnly pledge to actively champion diversity and counteract monopolies, duopolies, or oligopolies within the AI creative sphere. This commitment aims to foster innovation, fairness, and global collaboration.

ARTICLE V
SUPPORT FOR CULTURAL AND ACADEMIC ENDEAVORS

AI programs must be designed exclusively to support cultural and academic creators, refraining from autonomous generation of cultural and academic content.