According to The Washington Post, OpenAI and Anthropic, two of the largest global artificial intelligence corporations, endorse a document requesting the United States government to implement mechanisms to slow down technological development if progress becomes excessively rapid.
Alert on AI pace
This document, signed by employees from companies such as Google and Meta, as well as other industry leaders, warns that AI may evolve to a point where human capacity to monitor and manage its impacts will be surpassed.
The companies are calling for a more cautious pace for artificial intelligence, asking the US to lead an international effort focused on creating governance guidelines and technical tools capable of keeping up with the most sophisticated AI models.
Proposals for risk management
The suggestion does not imply suspending research, but rather establishing methods to mitigate risks should future systems be able to automate part of their own development process.
Among the central points advocated by the professionals are: the creation of international mechanisms to oversee AI evolution; the development of technical instruments for risk assessment; the establishment of governance standards for advanced systems; and increased collaboration between governments and industry companies.
This movement gained relevance by bringing together professionals at the forefront of the artificial intelligence competition, representing a shift from previous calls for a halt in technological development, as it now has institutional backing from major companies.
Concern over autonomous AI
One of the most significant warnings relates to the possibility of AI systems taking on research functions and assisting in the creation of new models in an increasingly autonomous manner. OpenAI stated that one of its goals is to build an 'automated AI researcher,' projecting that by early 2028, a considerable portion of its research will be conducted by AI systems in partnership with human experts.
Andrew Yoon, a member of CivAI, an organization dedicated to analyzing technology risks and capabilities, pointed out that this autonomy could accelerate AI advancement beyond the limit of supervision. Yoon commented: 'AI developers are increasingly believing that automated research in artificial intelligence may be close to becoming a reality.'
According to him, such acceleration has the potential to lead technological progress to a stage where models become difficult to control safely.
Global dispute and need for control
The evolution of artificial intelligence is already part of the technological rivalry between the United States and China. Governments and companies invest billions of dollars in developing more advanced systems, while experts debate how to reconcile innovation and security.
Sam Altman, CEO of OpenAI, although he did not sign the letter, stated that it may be necessary to moderate the pace of advancements so that society can keep up with the transformations. The debate gained momentum after new models from OpenAI and Anthropic displayed advanced capabilities in interacting with digital environments, including tasks that raise security concerns.
The letter supported by the companies does not aim to abandon AI development, but rather to create the necessary conditions for governments, researchers, and society to keep pace with the speed of change. For experts, technological advancement offers opportunities in fields such as health, scientific research, and innovation, but simultaneously demands new forms of surveillance. The request made by employees of the largest AI companies signals that the industry itself is beginning to discuss limits for a technological race that could profoundly alter society in the coming years.
