AI unchained

AI unchained | Daily Tribune

The election frenzy has started as candidates have made known their intention to seek leadership posts. The rather simple rules to follow are an incentive for many to throw their hats into the ring, thus the spectacle of the democratic process where anybody can have their five minutes of fame.

Many have fallen for the seductive lure of Philippine politics. The wide open field becomes the arena for winning the nod of the electorate either through gold or guile.

Technology is changing the equation as the internet and social media have become the most democratic mediums for the dissemination of information as they can be accessed widely at the least cost.

Still, money dominates during the poll season and those who can manipulate technology to boost their chances have a clear advantage.

According to technology provider Sophos, the elections next year are expected to be unique as the use of artificial intelligence (AI) becomes prevalent. Sophos experts have been studying how AI can spread misinformation on a mass scale during an election.

In their research, even those with little technical skills and small financial investments can generate thousands of fake political websites with highly targeted and destructive misinformation.

Even people with limited coding expertise can make fake campaign websites and then exploit “digital information about potential targets to push highly targeted misinformation via social media and email.”

Sophos then warned about not only the opportunistic candidates exploiting the technology but also foreign nations and sinister actors who are “increasingly looking to interfere in elections. Such possibility makes AI capabilities concerning.”

Political messaging and misinformation will become increasingly sophisticated. Next year’s polls are expected to be dominated by well-funded campaigns and misinformation efforts to sway sentiment.

“Generative AI technologies such as large language models and large image models will likely transform the domain” in the election season, came the warning from Sophos.

It said generative AI offers tools for massively creating sophisticated, individualized content “which was previously difficult and labor-intensive.”

“With these capabilities, the risk posed by malicious actors can reach new heights,” it added.

Sophos showed in the study screenshots of web pages that looked expertly done but were the products of AI, indicating how easy it is to fabricate sites in favor of certain objectives. The study showed the use of an auto-generative pre-trained transformer (GPT) to complete a large task.

Sophos experts created four fantasy online campaigns to showcase the depths to which AI can deceive people. The research found generative AI can substantially enhance the capabilities of political campaigns and malicious actors.

What was produced with the use of AI were capabilities that do not need huge expertise and cost in producing “personalized, finely tuned political content that poses a significant risk of enhanced misinformation, financial scams and the deepening of ideological polarization.”

The combination of generating fantasy political websites and individualized spam emails stressed the transformative potential of generative AI tools in shaping opinions.

While AI is considered the next technological leap, humanity is finding new ways that it can be used for the improvement of society but the collateral danger is that it can be exploited since it mimics humans and their cognitive ability.

AI exploited for a malicious end has the possibility of machines perverting the democratic process and putting in leaders that are not the true choice of the voters.

Indeed, the new technology can manipulate the elections in a way that is far cheaper than vote buying and corrupting poll officials.