Gen-AI vs opponents: Keep polls free of fraud
As India’s Lok Sabha polls approach, political parties have reportedly contacted digital content creators and podcasters to help get their messages across. Of course, major parties have IT cells on the job, too, but too random an outreach would enlarge the risk of losing integrity control.
Artificially generated videos of leaders have already made an electoral debut in this part of the world. Before Pakistan’s polls, AI-made speeches of its jailed ex-leader Imran Khan were reportedly going viral, with these clips said to be okayed by him.
But in Bangladesh’s elections, a deepfake surfaced of an opposition leader displaying ambivalence over the plight of Palestinians, a sure vote-loser in a Muslim-majority country. Indonesia’s poll lead-up featured an AI video of its long-gone president Suharto (1921–2008), miraculously resurrected by technology to guide voters.
It’s stunning how easy it has become to convincingly put words in the mouths of desk-spun politicians. Expecting AI tool-makers to police all fraud would be like trying to nail jelly to a wall, to use former US president Bill Clinton’s analogy for early attempts at internet control.
“Caveat elector" will have to apply. Voter beware.
. Read more on livemint.com
