Please register to access this content.
To continue viewing the content you love, please sign in or create a new account
Dismiss
This content is for our paying subscribers only
Explainer

AI’s quiet impact on 2024 campaign: 'Micro-targetting' explained

Large-language models are now more integrated in voter segmentation and targeting



AI in politics: Language models are now more integrated in voter segmentation, reshaping campaigns, reinforcing and tweaking old-school strategies.
Image Credit: Images were generated with deepai.org

AI has reshaped political campaigns dramatically. And in some ways, it did, representing a seismic shift.

Welcome to the era of AI-driven content-generation, messaging and "microtargetting". Recent studies shed light on the impact of AI and technology in US political campaigns.

Get exclusive content with Gulf News WhatsApp channel

Evolutionary

The impact of AI, and its influence in influencing public opinion, has been subtler than anticipated. It suggests that the era of merely photoshopping images is gone. Less of a revolution, and more precision.

In a word, AI has refined – rather than reinvented – political messaging, aiding campaigns with data-driven insights on individual voters. It suggests that the era of merely photoshopping images is gone.

Advertisement

While it's less of a revolution, it's more of precision, aiding old-school strategies, rather than overhauling them.

Studies

A number of studies bear this out.

#1. AI-Driven voter microtargetting

An MIT study found that AI-driven political microtargetting does work. It reveals how AI is used to identify detailed voter segments for targetted messaging. Companies like Resonate use AI to profile and target specific demographics, helping campaigns fine-tune messages for persuadable voters.

The technology enables campaigns to tailor strategies for groups identified by behavioural data.

In July, analyst Alexander Jones shared AI insights into a promising voter segment: “Cyber Crusaders,” a group marked by its blend of social conservatism and fiscal liberalism.

Advertisement
An MIT study published in June 2023 found targeting ads based on a single characteristic – like party affiliation – can increase persuasion “by up to 70 per cent” compared to using a universal ad.
Image Credit: Image Generator

Resonate’s models see them as a decisive demographic — young, religious, and inclined to trust online ads.

These insights are drawn from a vast dataset of around 250 million profiles, scrubbed for privacy and analysed to discern behaviour, attitudes, and possible voting shifts.

This advanced “microtargeting” has significantly changed how campaigns address undecided voters​.

How effective is it?

The study published in June 2023, and led by David Rand, an MIT professor, established what's been only hyphothesised up that point: targeting ads or messaging based on a single characteristic – like party affiliation – can increase persuasion “by up to 70 per cent” compared to using a universal ad.

Advertisement

It also challenged the idea that hyper-specific microtargeting is transformative in political campaigns, it found that adding layers of personal traits — such as ideology or age — does not boost effectiveness.

More important: it showed that while targeting in political ads is beneficial, the impact plateaus when additional audience attributes are factored in, signalling limits to microtargeting’s influence.

AI-driven micro-targeting
It is a marketing and political strategy that uses AI to analyse vast amounts of data, including demographics, behaviour, and personal preferences, to create highly personalised messages aimed at specific segments of an audience.

In political campaigns, AI-driven micro-targeting allows for the crafting of messages that resonate with individual voters or very narrow voter segments, enhancing the relevance and impact of campaign outreach efforts.

This approach is becoming increasingly common, especially on digital platforms, as AI can analyse patterns in large datasets and predict responses more accurately than traditional targeting methods.

#2. AI in social media ads, voter engagement

In a startling revelation from May 2023, the Republican National Committee (RNC) unleashed an AI-generated video targetting President Joe Biden. This marked a seismic shift in political campaigning.

It indicated the chaotic landscape where digital creations flood social media feeds based on carefully-defined, algorithm-driven parameters, shaping public perceptions in unprecedented ways.

Advertisement

'Wild West' of campaigns

Brookings has dubbed it the “Wild West” of campaign narratives, where the line between reality and fabrication blurs.

In a swift response to Biden’s reelection announcement, the RNC crafted a chilling vision of a dystopian future under Biden – illustrating how AI can rapidly distort reality and influence electoral outcomes. Amid this turmoil, Biden ultimately bowed out of the race, paving the way for Kamala Harris to take the lead.

US President Joe Biden was at the receiving end of AI-driven attacks when he announced he's running for a second term. Analysis shows that AI allows campaigns to craft highly targeted ads based on social media engagement patterns.
Image Credit:

This study highlights how campaigns use machine learning to design messages that resonate with different groups based on real-time social media responses, often leveraging neural networks and clustering to refine voter outreach.

By doing so, campaigns can respond to issues as they unfold, which helps them connect more effectively with constituents​.

Advertisement

#3. AI for predictive analytics in campaigns

AI has reshaped strategic decisions in campaigns through predictive modelling.

For instance, the Obama campaigns in 2008 and 2012 pioneered AI use in political campaigns, and since then, predictive analytics has become central.

The ability of AI to analyse voting behaviour and predict support patterns allows campaigns to allocate resources more strategically and optimise outreach efforts, especially for high-stakes races​.

#4. Fake news detection and fact-checking

An aspect of AI’s role in politics is in combating misinformation. AI models are deployed to monitor social media for false information and disinformation, helping to counteract misleading content.

These tools can flag potentially fake news stories and detect bots spreading misinformation. As noted in studies, this AI-driven monitoring became more prevalent after concerns that bots were influencing voters in previous elections​

Advertisement

AI social media monitoring tools use advanced algorithms to detect emerging trends and topics of discussion across various social media platforms.

By analysing patterns in user behaviour and content engagement, political campaigners (or businesses) can identify opportunities for content creation, tweak in messaging and political ad campaigns.

These studies underscore how AI's evolution is transforming traditional campaign practices, influencing everything from voter segmentation to combatting misinformation and improving strategic planning.

Potential benefits

The use of AI in these ways raises new ethical and practical questions for maintaining transparency – and fairness in political processes. Moreover, AI-driven sites by Logically.ai uses its AI resource to combat misinformation.

For all its possible downsides, AI has its upsides too. For example, it can potentially lower costs of AI-driven insights.

Potentially, this could narrow the resource gap between larger and smaller campaigns by making sophisticated targeting more affordable, helping smaller campaigns compete.

Refining campaign strategy, voter outreach

These AI models aren’t the dystopian overlords some imagine; they’re tools, woven into the fabric of human decision-making.

Like the printing press, radio, and television before it, AI is another step in how we communicate, gather information, and share knowledge.

As with every major technology, humanity isn’t replaced but transformed, adapting AI insights to enhance our collective intelligence.

While AI becomes part of our consciousness, amplifies our ability to understand and respond to each other in ever more sophisticated ways, human decision remains sovereign over our affairs.

For better or for worse, though no one knows for sure which way, and for how long it goes.

Sites for fact-checking claims

The resources, listed below, provide fact-checking to assess the truth of claims made in political statements, social media posts, and news coverage.

  • FactCheck.org - Monitors the factual accuracy of what is said by major U.S. political players in the form of TV ads, debates, speeches, interviews and news releases.
  • Politifact - Rates the accuracy of claims by elected officials and others who speak up in American politics.
  • Snopes.com - Conducts extensive fact-checking research on popular topics.
  • Washington Post Fact Checker - This Fact Checker column is "to 'truth squad' the statements of political figures regarding issues of great importance, be they national, international or local."
  • Duke Reporters' Lab - Maintains a database of global fact-checking sites. Use the map to explore fact-checking sites around the world.
  • Logically.ai - Uses artificial intelligence to combat misinformation.

Fact-checking academic sources

These resources provide updates regarding what publications have been recently retracted as well as a list of ethical standards for publications to follow.

  • Retraction Watch - Retraction Watch is a blog that reports on retractions of scientific papers. They offer an archive of these reports for scholars to stay up to date on untrustworthy academic publications.
  • SciCheck - FactCheck.org's SciCheck features focuses exclusively on false and misleading scientific claims that are made by partisans to influence public policy.
  • COPE (Committee on Publication Ethics) - COPE is committed to educating and supporting editors, publishers, and those involved in publication ethics, it aims to move the culture of publishing towards one where ethical practices are normalised.
Advertisement