US, UK, And EU Enter Into First Legal Treaty On AI Safety

The US, the UK, and the EU have all entered into a legally binding treaty called the Framework Convention on AI.

This is the first agreement that’s designed to ensure these regions follow proper AI safety requirements that align with human rights as well as democracy. The document lays down the foundation for important AI principles regarding AI systems.

This includes how user data can remain safe, how the law is obeyed, and how everything must remain transparent at all times. Every nation signing the agreement needs to adopt the best legislative measures and administrative tasks that are outlined in the agreement.

Therefore, you can think of it as a guide on how to best use AI and prevent its abuse.

Interestingly, the framework is not something new. It’s been under discussion since the start of 2019 and several countries are already following it including Israel, and Norway. Andorra. Moldova, Georgia, and San Marino.

In the past few months, we’ve seen a huge number of AI safety documents prevail. This just reflects how popular the technology is and what punishments signatories might incur when they go against it.

As mentioned by The Financial Times recently, this treaty is legally binding but following it can only be done through monitoring. This is never considered the best form of enforcing it.

Whatever the case may be, one thing is for sure. We feel it might serve as the most important blueprint for various nations that want to curate their own laws related to AI.

As we speak, America has several bills in the pipeline linked to AI while the EU has already passed historic AI regulations. Now, the UK is looking for its own framework on the subject.

California is the closest to passing its first AI Safety Act which many tech companies like OpenAI have tried to ignore for so long.

Highlighting the best AI standards and ensuring they’re met has always been a topic of great debate. Hence, it’s going to be interesting to see how signing this new treaty will change the way AI is used today.

As far as implementation of the treaty is concerned, that will come into play starting three months after several signatories formally approve it.

Image: DIW-Aigen

Read next: Despite Big Privacy Claims, Meta, Apple, and Google Share User Data with Law Enforcement Upon Request
by Dr. Hura Anwar via Digital Information World