Nowadays, the subjects of cybersecurity and AI are being put together more. The thing is that AI systems are usually under attack by bad actors, and that poses a major issue. Because of that, there’s a new bill attempting to aid in the security of AI systems.
We’ve been seeing more governmental bodies crack down on AI in general. The EU established the first laws surrounding AI safety, and the United States government has been bringing measures of its own. This means that more governmental bodies are realizing the potential danger of misused AI technology.
There’s a new AI security bill in the works
Named the Secure Artificial Intelligence Act, this is a bill that aims to help boost the security of AI systems in America. At this point, AI technology will be used in certain areas of governmental agencies, and this is why security is crucial.
What the act would do is basically boost research into what is called counter-AI. These are ways of manipulating AI systems for nefarious purposes. The two senators who introduced the bill, Mark Warner and Thom Tillis, actually classify these counter-AI methods under four categories. These are Invasion attacks, abuse attacks, data poisoning, and privacy-based attacks.
Along with research into AI measures, the bill will also require that the National Institute of Standards and Technology and the Cyber Security and Infrastructure Security Agency craft a large database of AI breaches. This will make it a lot easier to understand what types of cybersecurity incidents happened and keep track of them. They’re even required to include “near misses.”
When it comes to subjects like these, it’s always important to have bills that help boost security with AI systems. If there are any government agencies using AI, you can bet that there will be people out there trying to take advantage. So, steps like these are the best sort of defense.