REWRITTEN ARTICLE: AI IN POLITICAL CAMPAIGNS AND THE ROLE OF LANGUAGE MODELS IN MODERN-ELECTION STRATEGIES

No time to read?
Get a summary

Artificial intelligence (AI) and data science are gaining prominence across political landscapes. Parties recognize the leverage of technology, and campaign teams increasingly rely on tools like ChatGPT, Bing Chat, and Big Data to influence voters. This is evident in regional and municipal elections such as the upcoming 28M in the Community of Valencia and the Alicante province. Leading voices, including Andrés Pedreño, president of 1MillionBot, and Manuel Palomar, director of the Center for Artificial Intelligence-Cenid (UA, UMH and Diputación), advocate for language-generating technologies while cautioning about associated risks. They note how easily images or voices can be manipulated to mislead audiences, then spread from person to person via mobile apps like WhatsApp. In response to these dangers, they call for greater investment in education that equips all citizens with tools to resist manipulation.

The use of artificial intelligence is not new; the possibilities it unlocks now are far broader than in the past. Movements that began in major US universities helped seed streams of ideas through social networks, a tactic that influenced national politics and autonomy. “There has been a democratization of knowledge. It started with advanced statistics and then moved on to artificial intelligence. This made it possible to launch ideation panels via networks like Facebook or Twitter,” explains Pedreño.

The former rector of the UA cautions against risks such as miscalculation by various actors who create streams of opinion to benefit themselves, sometimes deriving harsh criticisms or applause based on context. “AI is used to identify trends and critical ideas that can help guide voting. Advanced statistical tools were effective, but AI capabilities are far higher,” Pedreño warns, before addressing the dangers: “The most important tool against manipulation is education. Freedom of opinion, information and press is essential for voting decisions, and a well-educated population is critical. European governments should invest in education so citizens can verify the news.”

language production

The discourse from Palomar centers on language models such as ChatGPT or BingGPT. “Just as technology supports progress in any field, it offers numerous political uses, including drafting documents, delivering speeches, broadcasting on networks, monitoring trends, and supporting the development of cultural, educational or tourist programs. These tools have established a lasting presence and now touch all areas of knowledge,” states the Cenid leader in this analysis.

Research indicates that AI can accurately predict voting preferences, a finding of great importance as campaigns craft increasingly personalized messages. Within selection processes, bots and social network profiles managed by algorithms may attempt to sway votes toward their own side, a dynamic that played out during regional elections and may align with the left-right spectrum in 28M campaigns.

biggest innovation

Experts agree that the most disruptive development is the emergence of ChatGPT, a chatbot capable of answering almost any question in natural language. While still evolving, its potential appears vast. For now, parties approach its use cautiously, mindful of boundaries. An example is the Unides Podem alliance, which has stated it will not use ChatGPT immediately but recognizes its future potential to assist in preparing speeches, discussions or program content.

The axes of the campaign: housing, taxes, inflation, climate change and water

The central question is whether an AI language model-driven chat can meaningfully influence election outcomes and become a cornerstone of modern political communication. Available tools, paired with fresh strategies, may augment and accelerate political campaigns. This shift has contributed to the evolution from large rallies toward more targeted micromarketing strategies.

Compared with traditional social networks, ChatGPT offers capabilities for generating complete content, improving management, and boosting teamwork. The benefits for political communication include auto-answering questions and messages, reaching larger audiences, automating content creation with personalized responses, and enabling real-time conversations, data analysis, and generation of reports and statistics.

Artificial intelligence

The scope of AI is so broad that ad creation could even move into televised spaces, a trend national parties have begun exploring. For example, En Comú Podem has produced an advertising spot that may convey dual messages to Barcelona’s municipal authorities. Experts like Pedreño and Palomar argue that AI offers a wide range of capabilities to improve people’s lives, but it must be regulated and cannot replace human decision-making.

With greater uptime, big data and social-network analysis have driven success stories in earlier campaigns, where traditional voting intent surveys via phone calls or interviews were bypassed. Data science now allows near real-time analysis and cheaper methods to gauge voter sentiment. The same techniques enable rapid analysis of thousands of tweets and blog posts to assess alignment with the goals of the organizations commissioning them.

Beyond these technologies, social networks remain influential, with TikTok as a standout platform. Parties have sought to engage younger voters by establishing profiles and delivering messages in a way that resonates with a younger audience, especially as the 28th approaches. Regional and national campaign teams have encouraged exploration of TikTok and BeReal, among others, as part of a broader social-media strategy. The PSOE, for instance, includes these options in its 28M campaign guide, alongside Facebook, Twitter, Instagram and Twitch to reach audiences spontaneously.

Since 2011, the General Election Regime Organization Law regulates what can be done on social networks, restricting propaganda to a fifteen-campaign-day window. While the rules are not perfectly clear, election authorities intervene as needed to adapt to evolving practices.

PSPV-PSOE

In PSPV-PSOE, AI tools speed up dissemination and streamline communication tasks by converting audio to text. Big Data helps segment messages and identify which tweets or Instagram posts had the greatest impact. On social networks, PSPV uses them all, from Facebook to BeReal, including Twitter, Instagram and Twitch.

PEOPLE’S PARTY

The party analyzes each user’s digital experience to identify real problems and propose concrete solutions. This is the primary use of technology for the campaign. For the Consell candidate Carlos Mazón, the focus is on portraying his authentic image, with a dedicated data analysis team evaluating social-media performance.

CITIZENS

Ciudadanos relies on programmatic advertising to deliver messages where they matter most. Libertarians deploy data-driven tools and algorithms to tailor messages to users based on background and behavior. Oranges tracks tweet content to detect topics of interest.

COMMITMENTS

Compromís emphasizes targeting messages to specific demographic groups, such as youth or seniors, and uses social-network advertising plus podcasts. The campaign includes radio-ready formats of proposals to reach the entire community.

VOX

The Vox campaign team analyzes attendance-based data by district and population, then feeds it into AI-based estimations. This enables focused visits to certain municipalities and tailored advertising. They also measure abstention as part of the planning.

UNITS WE CAN ACHIEVE

Within the left alliance, pilots include ChatGPT and Bing for drafting content. Yet Unides Podem notes that mastery of these tools is still in early stages and not yet deployed broadly. They believe refined usage could prove valuable for interviews, speeches, debates, and program development when the time is right.

No time to read?
Get a summary
Previous Article

Pixel Fold: Google’s First Foldable Smartphone with Dual Displays

Next Article

When was the last time an Italian team played in the UEFA Champions League final?