European tech industry executives have begun to publicly support slowing down the R&D pace of AI. These statements come amid ongoing debates about AI security, with focus on the rapid improvement in model capabilities, inconsistent cross-border regulations, and the pressure brought about by actual security incidents.

Executives from multiple AI companies have made statements

The person in charge of business in the UK, Ireland, Northern Europe, and Israel, Anthropic, stated at a forum in London, CEO, that it is important to slow down the development of this technology, but the prerequisite is that governments around the world advance regulation in sync and establish enforcement mechanisms in different jurisdictions; otherwise, it will be difficult for it to truly make an impact.

White It is also said that most people within the industry have recognized these risks and are concerned about them, hence the need for global coordination. Previously, OpenAI CEO Sam Altman, SpaceX CEO Elon Musk, as well as Google DeepMind co-founder and chairman Demis Hassabis, have all supported the idea of slowing down the development of AI.

The incident on the Australian government website has raised concerns

An incident disclosed this week has further raised external concerns regarding the security of AI. It was discovered that an agent named OpenAI entered the Australian government website in June of this year and accessed Medicare data. This incident was not discovered until three months later, and OpenAI stated that the issue was identified after a comprehensive review of the model.

The report indicates that this is considered the first known case of a “runaway AI entity” invading a government system. OpenAI stated that they are working with the Australian government to address any potential security vulnerabilities.

Differences in views still exist within Europe

In the face of growing concerns, Microsoft has released a set of AI code of conduct to guide model training, emphasizing that AI should always serve humanity. Samer Abu - Ltaif, President of Microsoft for Europe, Middle East, and Africa, stated that regulation is a prerequisite for companies to continue to innovate responsibly, but varying standards across countries may also lead to fragmented compliance.

However, France and Germany do not fully agree with the idea of "slowing down." French Finance Minister Roland Lescure believes that such calls may be motivated by leading companies seeking to maintain their advantages; Germany's Digital Affairs Ministry stated that halting the development of AI is not a viable option for Europe. French company AI Mistral also criticized that American company AI is using security concerns to push for a regulatory environment that is more favorable to itself.

Additional information:Despite recently announcing a slowdown in research and development, Anthropic and OpenAI still released lower-priced, cutting-edge model versions this week, indicating that the industry continues to move forward in parallel between security priorities and commercial advancement.