By Necva Tastan
ISTANBUL (AA) - In December, Pakistan’s former Prime Minister Imran Khan delivered an impassioned speech at a virtual rally of his Pakistan Tehreek-e-Insaf party.
The only catch: Khan was actually behind bars at the time, where has been since last August after being convicted in a graft case.
The speech was an audio clip featuring a voice clone of Khan generated using artificial intelligence, marking a first in the country’s politics.
Over in the US, Shamaine Daniels, a Democrat running for Congress in Pennsylvania, has employed Ashley, an AI-powered robocaller touted by her creators as the first political phone banker.
These are just some of the instances illustrating the growing footprint of AI technologies in political strategies.
Bruce Schneier, a US-based security technologist and author, believes these cases could mark a potential shift in the political realm, adding that it would “not be long before AI develops campaign strategies and suggests what positions you should take.”
However, he is uncertain about the extent or nature of these changes.
“I think this technology will both help and hurt,” he told Anadolu.
On the one hand, there are cases where we have AI systems “helping sort of regular people become candidates, which seems positive,” he said.
“Right now, political strategies rely heavily on political exalts … on human experts who tell the candidates what to do, what to say, and how to say it,” he said.
“One thing (AI does) is it scales. So local candidates who couldn’t afford to hire a political consultant can use an AI. That would be a difference because it’s cheaper.”
On the other side, there is a “real risk of demagoguery,” he warned.
“That is a potential problem if there’s no accountability, there’s no auditing what the candidate chatbot is saying, or what the fundraising telephone bot is saying,” he explained.
- The fake factor
Earlier this month, there was a robocall with an AI voice clone of US President Joe Biden, where he was discouraging New Hampshire residents from voting in the state’s presidential primary.
The incident has since sparked a new wave of debate over the use – rather misuse – of AI technologies for political purposes.
In Schneier’s view, the threat of AI in this particular area still needs to be assessed and evaluated.
“We’ve been seeing fake videos, fake videos, fake type of misinformation for a good decade now … We’ve already had fake everything affecting public opinion for a good decade,” he said.
It is also “unclear how effective these things are moving public opinion,” he added, citing a lack of concrete data.
On disinformation campaigns, he said there are already other tools being used for that purpose, so the difference that AI makes in that is still uncertain.
Commenting on the global push for control over AI technologies, Schneier emphasized that “regulation is not keeping pace with the technology.”
The EU’s AI Act remains the only major regulation in the field, he said, referring to the law that European officials agreed on in December.
The act, however, still has to go through more approvals and is not expected to take effect until at least 2025.
“The EU AI Act … it’s been a few years in writing, so is it already obsolete?” said Schenier.
“Things are changing so fast … What was true in the summer is not true today.”