As the 2024 election season unfolds, battleground states are experiencing an influx of contentious campaigning by politicians, political action committees and, thanks to social media platforms, the electorate. Messaging continues to follow in the rich tradition of brief-but-impactful slogans and soundbites, now made accessible to thousands of voters online. But in their quest for robust bang-for-your-buck returns, some campaigns and strategists have embraced questionable deepfake technology, aimed to deceive a weary public already inundated with disinformation. Enhanced education and an awareness are needed.
Allegations are already flying and discord is being sown. On Sunday, former President Donald Trump claimed a photograph showing a large crowd at Vice President Kamala Harris’ campaign rally in Detroit was AI-generated. It was not. That a leading presidential candidate deemed it appropriate to air such a bogus assertion, and calculated that his constituents would find it credible, is a sign of the times.
Hawaii during the recent primary election was fortunate to remain largely unscathed by the deepfake threat — local contests are relatively civil, if not friendly, affairs. But with a presidency and control of both chambers of Congress on the line come November, island voters will be increasingly exposed to national-level campaign tacticians able and willing to deploy media, manipulated or generated by artificial intelligence tools, with an express impetus to deceive.
During the past legislative session, state representatives sagaciously determined deepfakes a threat and took preliminary steps to erect a bulwark of safeguards designed to protect Hawaii’s voters. House Bill 1766 proposed moderating the flow of “synthetic media” messaging in advertisements, while Senate Bill 2687 would create a petty misdemeanor for “distributing, or entering into an agreement with another person to distribute, materially deceptive media” without a disclaimer. HB 1766 died in committee, but SB 2687 became Act 191 in July and now serves as groundwork on which future false media regulations can be built.
Still, the wide-reaching and immediate nature of social media, where many deepfakes are shared and proliferated, is not bound by local law, and as the general election nears, Hawaii faces a voter education challenge not seen in past races. Traditional media outlets have proven unable to keep pace with the deluge of disinformation and misinformation surrounding the presidential race, and at times amplify unscrupulous posts with overwrought fact-checking.
Content policing by social media networks has been less than effective. In late July, X owner Elon Musk recirculated deepfake audio of Harris. The package was ostensibly crafted as parody, which X allows per its content policy, but is presented minus the usual “manipulated media” disclosures seen on other platforms like those owned by Meta.
There is only so much X, Meta and others can do to inform and protect users from deepfakes — if they choose to do anything at all. A few suggestions when encountering questionable media online:
>> Research — seek out reporting from established media outlets, some of which have access to specialized AI detection tools.
>> Cross reference — check other platforms or networks to corroborate a post’s validity.
>> Consider the source — where did the media originate, who posted it and where was it disseminated?
>> Rely on your peers — comments can reveal a trove of information from experts, eyewitnesses and insiders.
Ultimately, voters bear the onus of discerning falsehood from fact. It’s a particularly daunting task given that rapid development in the AI field, and commensurate advances in synthetic images and audio, make it difficult to distinguish real from fake. But awareness is half the battle.
