Margrethe Vestager, Vice President of the European Commission, said we need to be pragmatic when it comes to fine-tuning rules around AI as it continues to develop quickly.
What is the Complete Story?
According to the AI Act’s current draft, it seeks to categorize Artificial Intelligence into levels of risk to customers. And spam filters and AI-enabled video games fall into the lowest risk category.
However, evaluating credit scores or access to loans and housing falls into high-risk AI systems. This is where tech should put their strict controls on. But as Artificial Intelligence continues to develop rapidly, Margrethe Vestager made a statement that when it comes to fine-tuning rules around the technology, it becomes necessary to be pragmatic.
She added 80% now, then 100% never, so let’s get started and then return when we learn and correct with others. There is definitely a risk that AI will influence upcoming elections. In this case, the challenge for intelligence and police services should be on top.
So that where there is a risk, they get ahead in the race, utilizing the technology. However, the risk of being manipulated is enormous if your social media handle is scanned. We will undermine our society entirely if we get into a situation where we believe nothing.
In March, many tech leaders signed a letter calling AI poses extinction of humanity risk and for a pause in developing Artificial Intelligence more powerful than ChatGPT-4. But this was not realistic, as Margrethe said discrimination is a bigger concern than human extinction from AI.
Restrict the Use of Biometric Identification Systems
The Artificial Intelligence Act seeks to restrict the use of biometric identification systems and indiscriminate data collection from CCTV footage or social media for facial recognition. We have to set some strict rules so that it is not used in real-time.
But this should be at a specific moment when you’re looking for a missing child or a terrorist is fleeing. The world’s first rulebook on the development of AI has to agree on its final version before the AI act finalizes.
It is not expected to come into action before 2025.