18 countries have formed a fully non-binding consensus for the development of safe artificial intelligence technology. This agreement creates the framework for what should be done for safe artificial intelligence.
Artificial intelligence-focused technologies have entered our lives in many areas. Now we can create images and even videos with artificial intelligence, can write articlesWe can develop applications. This being the case, many segments of society believe that artificial intelligence technologies may pose a risk believes. This is not a wrong approach, because an artificial intelligence that can write articles can very well create a scenario.
Discussions like these are the reasons why countries should focus on artificial intelligence. to develop laws It led them to take steps. Even the USA took the first steps towards this. The European Union has also been working on an artificial intelligence law for a while. Now, including developed countries such as the USA and the UK, 18 countriessigned a non-binding memorandum. This consensus explains what safe artificial intelligence technologies should be like.
What’s in the agreement?
In addition to the UK and the US GermanyThe agreement, signed by countries such as Italy, Estonia, Poland, South Korea, Czechia, Singapore, Nigeria, Israel and Australia, in its most general terms, protects the consumer from the possible risks of artificial intelligence technologies. to be protected aims. The agreement, which is the first international agreement, will ensure that user data is protected and the developed systems are resistant to hackers. give advice is available.
According to the consensus, a safe artificial intelligence technology should pay attention to the following four items:
- Safe design.
- Secure development.
- Secure distribution.
- Safe operation.
Now let’s look at these items on how to carry out Let’s look at the information.
“Safe design”
- About threats and risks increasing staff awareness.
- For the artificial intelligence system to be developed of threats detected and modelled.
- Functionality, performance and security features of the system combination of creation.
- The security of the artificial intelligence model suitable for the system to be developed to highlight.
RELATED NEWS
The USA Took the First Concrete Step for the Artificial Intelligence Law: The Legal Grounds That Will Be Required to Be Abided by Have Been Officially Established!
“Secure development”
- System stakeholders to be secured.
- All assets owned by the artificial intelligence system to be developed definition and protection.
- Data, models and claims archiving.
- risk assessment to be done.
“Secure distribution”
- infrastructure to be secured.
- artificial intelligence model protectionupdated regularly.
- For possible problems establishing procedures.
- Artificial intelligence technology to be developed assuming responsibility.
- User friendly preparation of interfaces.
“Safe business”
- System behavior Following.
- System logins monitoring.
- Trustworthy updates presenting.
- The artificial intelligence system put into operation collecting the results. Sharing when necessary.
All of these will ensure the provision of safe artificial intelligence technologies for these 18 countries. General considerations. However, as we mentioned at the beginning of our news, all of these are not binding. If you want to reach the full consensus here You can use the link.