New Delhi: Meta CEO Mark Zuckerberg has pushed back against calls for a broad slowdown in artificial intelligence development, arguing that individual AI companies have enough reasons to build and test their systems safely.
Zuckerberg’s comments come as technology companies face growing debate over the pace of AI development and the risks associated with increasingly capable models. The discussion has intensified as companies invest billions of dollars in computing infrastructure and compete to develop more advanced AI systems.
Zuckerberg has argued that AI companies can manage safety on their own rather than relying on an industry-wide agreement to slow development. According to his comments, companies have strong incentives to ensure that their products are safe because of customer expectations, competition and potential legal risks.
Meta has itself delayed the release of some AI products while carrying out additional safety work. Zuckerberg pointed to the company’s decision to delay its Muse AI agent as an example of taking more time to address security concerns. Meta has also used independent evaluators in parts of its AI development process.
His position differs from that of some other technology executives. Anthropic CEO Dario Amodei has called for greater caution around the development of advanced AI systems and has supported discussions around managing the speed of development. OpenAI CEO Sam Altman has also spoken about the need to pace frontier AI development and use independent evaluations for advanced systems.
The disagreement is not about whether AI safety matters, but about how the industry should manage it. Zuckerberg’s approach places greater emphasis on individual companies making their own decisions based on the systems they are developing. Other industry leaders have argued for stronger coordination and common safeguards.
The issue has become more important as AI systems are increasingly being used for coding, research, business operations and other complex tasks. The next generation of models is expected to become more capable, increasing both their potential economic value and the importance of testing them before wider release.
Meta has also said that much of its AI computing capacity is being used to develop products for existing users rather than being focused entirely on systems capable of autonomously improving themselves.
The debate is likely to continue as technology companies invest heavily in AI infrastructure and model development. Regulators, researchers and businesses are also examining questions around security, transparency and responsibility.
For the industry, the central question is whether AI safety can be effectively managed through company-level testing and safeguards or whether common standards and coordinated rules will eventually be needed as AI systems become more powerful.
