UNITED States President Donald Trump recently held a meeting with the chief executives of leading technology companies to address public concerns about artificial intelligence (AI) agents going rogue and hacking confidential data of companies and governments. At the end of the said meeting, it was agreed that AI companies would self-police to prevent potential dangers.
I am not surprised that the companies immediately agreed to devise a self-policing regulatory framework because they have invested around $1 trillion already and do not want government regulations.
Trump, too, does not at all appear to be in favour of any government regulation because he thinks China would then win the crucial AI race. So, clearly, there isa basic conflict of interest in the matter.
In essence,self-regulation doeshave an inherent weakness. AI companies are competing for development to get more market share, and in this race they may ignore risk and caution. Recently, an OpenAI agent hacked healthcare records of Australia. The dangers posed by AI, if gone rogue, can cause a tremendous amount of disruption to global health, commerce, communication and banking systems.
For instance, if AI agents hack data of a country’s banks, the entire banking mechanism in that country would be disrupted, resulting in massive losses.
Similarly, if AI hacks data of a busy shipping port, it will disrupt movement of containers, resulting in shortage of goods in the country.
In view of such apprehensions, AI should be regulated by governments just like pharmaceutical and other industries are regulated for everyone’s safety. AI must be faster, but it must be safer at the same time. Experience shows that regulations always help companies to grow better.
After the financial crisis of 2008, the US tightened bank regulations and American banks are stronger and more prosperous today than they were before that time.
Ejaz Ahmad Magoon Dubai, UAE