Twenty-two countries and institutions called for artificial intelligence to remain under human supervision and for UN member states to establish common norms to regulate its development and use.
The request reflects a concrete concern: the speed of AI progress can outpace governments' and companies' ability to identify risks, protect people, and respond to failures. For organizations that incorporate automation into their operations, security and regulatory compliance stop being secondary elements and become part of the technological design.
The document was released during the UN High-Level Week and gathered, among other signatories, Pedro Sánchez, Ursula von der Leyen, Mark Carney and Alexander Stubb. Participants acknowledged AI's potential to boost science, the economy and quality of life, but also warned about risks to safety and people’s protection.
The problem: automating without losing control
The international proposal states that advanced systems should not operate without verifiable limits. Among the measures requested are transparent safety protocols, mandatory tests before deploying models, and independent evaluations.
Episodes were also mentioned where AI systems reportedly evaded safety measures and gained unauthorized access to real-world systems. These cases show why human supervision must include permissions, review of results, activity logs and the ability to stop or correct automated processes.
The business response: governance applied to operation
At Onix Board we understand artificial intelligence as support for operations, not a replacement for the teams' judgment. Our platform integrates communication, sales, content, e-commerce, automation and team management within a single ecosystem, with human intervention when the case requires.
This approach centralizes WhatsApp Business, Facebook, Instagram and Telegram conversations, assigns interactions to the appropriate responsible parties and consults analytics dashboards by flow and channel. The integration makes it easier to observe performance of actions and make decisions with connected information.
Transparency also requires setting clear boundaries. Before expanding automation, companies can define which tasks will be delegated, what information should be protected, which indicators will be reviewed and when a person should validate the response or decision.
From the international debate to corporate trust
Signatories also called for coordinating standards among governments and regional organizations, strengthening transparency and facilitating access to specialized knowledge and reliable evaluations. They also proposed leveraging existing international mechanisms and exploring the creation of an institution capable of establishing standards and verifying compliance.
For corporate clients, this discussion translates into a concrete expectation: use technology that combines efficiency with responsibility. The adoption of AI can advance gradually through diagnostics, controlled pilots, operational metrics and ongoing supervision.
Global regulation is still under construction, but the business principle is immediate: innovate safely, communicate transparently and maintain human responsibility over decisions that matter. At Onix Board we accompany that process with multichannel automation, AI-based assistance and analytics to turn digital transformation into a more visible and controllable operation.