The US administration published a commitment titled Joint Commitment on Frontier Responsibilities for artificial intelligence at the meeting held with executives of major technology companies. The issue of accountability, which previously came to the fore when discussing changes in OpenAI’s security management, is this time addressed with the phrase “superintelligence”.
Executives such as Mark Zuckerberg, Jeff Bezos, Elon Musk and Dario Amodei attended the meeting. The US administration has highlighted the name “superintelligence”, which establishes a more ambitious framework, instead of “artificial intelligence”.
Document; It includes topics such as security of advanced systems, sharing of risks and responsibility to the public. However, the commitment is not legally binding. Donald Trump called the document “morally binding.”
This distinction is important because the document does not contain a clear enforcement mechanism if signatories fail to comply with the rules. It has not yet been detailed how companies’ own tests and disclosures will be audited.
Implementation will show the impact of the commitment
Common text could create a minimum liability language between companies. However, unless independent auditing, incident reporting time and measurable security thresholds are defined, it will be difficult to compare the impact of the document.
In addition, the use of the name “superintelligence” combines the limits of today’s systems and future goals in the same discussion. Therefore, it is necessary to evaluate public communication and technical security measures separately. The next step will be whether general principles will translate into concrete and controllable rules.