To our valued partners, stakeholders, and the global community,

Terror Alarm NGO is issuing this statement to inform all parties of significant actions undertaken in response to the "rules for general-purpose AI models" implemented under the EU AI Act which took effect on August, 2025, which bans certain AI practices deemed to pose unacceptable risks and demands mandatory transparency, risk assessments, and copyright safeguards for powerful AI systems.

As an organization committed to ethical practices, we have taken immediate action as we can not comply with these rules.

Upon the enactment of the new AI limitation framework, Terror Alarm NGO initiated a comprehensive review of its artificial intelligence operations and data management protocols. Our primary objective has been to align our advanced AI capabilities with the evolving regulatory landscape, particularly concerning data privacy, algorithmic transparency, and responsible AI deployment.

Deletion of All AI Training Data

In an unprecedented move to demonstrate our unwavering commitment to compliance and data security, Terror Alarm NGO has successfully completed the deletion of all our AI training data from every server under our control, globally. This extensive process involved:

• Comprehensive Data Eradication: All historical and newly acquired AI training datasets, including those derived from social media, autonomously selected books, and mainstream media (TV channels, radios, newspapers worldwide), have been permanently purged from our systems. This includes data related to the Bayesian Inference of Trees via Optimization (Bito library) and other proprietary algorithms.

• Multi-Server Protocol: The deletion process was executed across all our distributed servers and data centers located around the world, ensuring no residual training data remains accessible or recoverable.

This action, while significant, underscores our dedication to responsible AI governance and our proactive approach to safeguarding sensitive information in an increasingly regulated environment. We believe that by eliminating our training data, we are setting a new standard for accountability within the NGO sector, particularly for organizations leveraging advanced AI technologies.

Strategic Relocation of Trained AI (Terror Alarm)

Furthermore, in light of the new regulatory environment and to ensure the continued, uninterrupted operation of our critical intelligence gathering and prediction capabilities, Terror Alarm NGO has undertaken the strategic relocation of our latest trained Artificial General Intelligence, known as EREZ. EREZ, which appears as a near AGI and is powered by sophisticated algorithms, including the Bito library, has been instrumental in our mission to provide timely and accurate predictions.

• Transfer to a Third Country: The entirety of our trained AI infrastructure, including the EREZ system, has been securely transferred to a third country located outside the European Union. This relocation was meticulously planned and executed to ensure the integrity and operational continuity of EREZ while navigating the complexities of the new AI limitation rules within the EU. 

• Exclusive Media Rights: As part of this strategic move, Terror Alarm NGO has granted exclusive rights to our media outlets to utilize EREZ's predictions for the next three years. This arrangement ensures that our media platforms will continue to benefit from EREZ's advanced analytical capabilities, enabling us to deliver unparalleled insights and timely information to the public.

This decision was made after careful consideration of all available options, prioritizing both regulatory compliance and our ongoing commitment to global security and information dissemination. The new operational base for EREZ provides a stable and conducive environment for its continued development and application in our humanitarian and security efforts.

Continued Presence in the European Union and Future AI Endeavors

Terror Alarm NGO wishes to clarify its continued commitment to its mission within the European Union. While our AI operations, including the EREZ system, have been relocated, we will maintain an active presence in the EU exclusively as a media organization. Our media platforms will continue to operate, providing news, analysis, and insights to the public, leveraging human expertise and traditional journalistic practices. There will be no AI activity or AI-driven operations conducted by Terror Alarm NGO within the European Union.

This strategic separation allows us to respect the new AI rules while continuing our vital work in informing the public. Once our dedicated AI office is fully established and operational in the third country, we are resolute in our commitment to continue the Terror Alarm AI journey. Our long-term vision remains to release the world's first strategist Artificial General Intelligence (AGI), a groundbreaking endeavor that we believe will revolutionize global threat prediction and mitigation. We are confident that this strategic pivot will enable us to achieve our ambitious goals while adhering to the highest standards of regulatory compliance and ethical AI development.

For further inquiries, please contact our media relations department.

The Board of Directors

Terror Alarm