AI and social media: the new architecture of political communication

In modern politics, two transformational forces are reshaping how candidates communicate with voters: social media and artificial intelligence (AI). Together, they have disrupted the traditional campaign playbook of press releases, televised debates and newspaper endorsements. Candidates now speak directly to voters, shape narratives in real time and build communities that thrive both online and on the ground.

From Zohran Mamdani’s viral-savvy mayoral campaign in New York to the rapid adoption of generative AI in campaign operations, we are witnessing a historic shift in the mechanics and expectations of political communication.

In this blog, the GRA team examines how social media is rewriting the rules of message delivery, and how AI is reinventing the way campaigns craft, target and scale their engagement with voters.

Social media: from broadcast to community & culture

The widespread use of social media today allows political candidates to engage with voters directly, craft narrative moments, mobilise volunteers and drive participation in ways that broadcast media cannot. The Zohran Mamdani campaign is a perfect illustration of this. Starting from polling at one percent, his team leveraged highly produced videos alongside every day clips filmed on his phone, generating content that young voters shared, remixed and rallied around. Examples ranged from a polar‐plunge video to highlight housing issues to a stroll across Manhattan on a Citi Bike as a symbol of connection and accessibility.

What this demonstrates is a fundamental change in how candidates communicate:

• Authenticity & storytelling: voters are increasingly suspicious of overly crafted campaign videos, they respond when a candidate feels real, present and imperfect. Mamdani recognised this: “Too much of politics has become artificial, has become the creation of a self that is actually divorced from the way in which you grew up in the world” (Wired).
• Platform savvy & participation: Mamdani’s team not only produced highly polished videos, but also iPhone-style clips shot on-the-go. He invited his fellow New Yorkers and supporters to participate in designing stickers and joining scavenger hunts. Much of the content about the campaign didn’t come from the campaign at all, it came from Mamdani’s supporters. Fancams, fan art and videos were created not in collaboration with Mamdani but for him and his campaign.
• Community building: According to Mamdani the campaign had grown to more than 85,000 volunteers, many of whom first engaged via a video, tweet or photo. Mamdani used social media to connect with New York’s diverse communities by sharing localised, relatable content, from multilingual videos to neighbourhood-specific stories and everyday city experiences. New Yorkers saw themselves reflected in his campaign and engaged directly online.

Mamdani’s campaign is a clear marker of how campaign communication is evolving: from one-way broadcast (TV spots, print ads, speaking at events) to multi-way, digital engagement where candidates build movements online, combine culture and policy and activate community.

AI: the new engine of political campaigns driven by precision, personalisation and speed

AI-enabled analytics now allow campaigns to identify highly granular voter segments – distinguishing, for instance, between TikTok-native younger voters, Facebook-using older voters and language-minority communities – each receiving tailored tone, content and calls to action (LSE).

Alongside this targeting, AI dramatically accelerates content production. Tools can generate written messages, visuals and videos, and even power interactive ‘spokesbots’ – as seen in Mexico City, where an AI representative was deployed to engage residents directly (Open Data Institute).

The result is a campaign cycle that moves at unprecedented speed. Tasks that once took days – scripting, editing, publishing – can be completed overnight. Messages shift in real time to match events, exploit social-media moments and optimise for virality.

In practice, this means:

• Precision messaging & micro-targeting: AI allows campaigns to segment voters with greater granularity, tailor messages to their preferences/behaviours, and deliver those messages via the platforms they use.
• Content generation at scale: AI accelerates content creation (email drafts, social-posts, ad copy, even automated chatbots). This increases speed and reach but raises questions about human authenticity.
• Real-time engagement & iteration: AI enables faster testing of messages (what works/doesn’t), quicker adaption of creative and more dynamic use of social media campaigns.

Together, social media and AI have transformed political campaigning, changing the way voters are reached, engaged and mobilised.

Changing the rules: what’s different now

One of the most significant changes is the ability of candidates to bypass traditional media and communicate directly with their voters – shaping their own narrative and mobilising supporters in real time. Mamdani’s campaign exemplified this: social platforms built momentum, activated volunteers and generated mainstream media attention as a natural outcome rather than a prerequisite (Wired).

Mass messaging has also been replaced by hyper-personalised outreach. AI-driven analytics allow campaigns to tailor content to distinct communities, speaking to each group in ways that reflect their values and concerns. Campaigns are increasingly becoming culture engines, embracing remix culture, online fandom, and participatory creativity – from sticker contests to viral challenges – turning supporters into active co-creators (Guardian).

Speed and experimentation now define the political communication environment. Platforms evolve quickly, conversations shift in minutes and AI tools enable campaigns to produce, test and refine messages almost instantly.

Finally, the boundary between campaigning and governance is blurring. Mamdani has reflected on how social media should evolve once in office: not as escapism, but as a way to “tell the stories of the world and then transform it” (Wired). This shift demands that transparency and authenticity continue beyond election day and into public service.

Risks and ethical stakes

The integration of social media and AI brings ethical challenges that can threaten both candidate credibility and the integrity of democratic processes. A core tension lies between authenticity and automation: voters respond to real human voices, yet increasing reliance on AI-generated content risks diluting that connection. As Mamdani notes, politics often becomes artificial when candidates present a polished ‘snapshot’ version of themselves (Wired).

There are also serious data privacy and targeting concerns. AI-driven micro-segmentation raises questions about how voter data is collected, analysed and used: Are people aware they are being targeted? Are certain groups seeing different messages in ways that reduce transparency? At the same time, generative AI exponentially accelerates the ability to generate misleading or false but convincing content, heightening the risk of disinformation, as highlighted by the Reuters Institute.

These challenges are intensified by regulatory gaps. Many governments are still defining how to oversee AI in political contexts, and analysts at Media Laws stress the need for clearer legal frameworks – especially during elections – to safeguard accountability and voter rights. Beyond campaigns, governance risks emerge if AI systems are adopted in official government communications; without strong transparency and oversight, public trust could erode.

A practical guide for modern campaigns

If campaigns use digital and AI tools thoughtfully and responsibly, the following principles offer a reliable best-practice blueprint:

• Build digital strategy from day one: use social media for genuine connection, mixing polished content with real, everyday moments (as seen in Mamdani’s approach).
• Create communities, not audiences: encourage participation through co-creation, contests, and grassroots engagement to turn supporters into collaborators.
• Use AI as an enhancer, not a substitute: apply AI for data analysis, segmentation and content generation while keeping human judgment and voice at the centre.
• Iterate in real time: monitor online discourse, test what resonate and adapt quickly, while grounding messages in lived experience.
• Prioritise ethics and transparency: be open about how data and AI tools are used, avoid manipulative tactics, and protect voter trust.
• Extend digital engagement beyond elections: integrate social media and AI into governance to strengthen accountability, accessibility and long-term community connection.

Conclusion

We now live in a political landscape shaped by technology. Social media and AI are no longer peripheral tools; they are central to how campaigns communicate, organise and build power. Zohran Mamdani’s campaign shows how social platforms can mobilise communities, activate supporters and bypass traditional gatekeepers to create genuine momentum. At the same time, AI offers unprecedented capabilities for targeting, scaling and personalising communication – while introducing serious questions about ethics, authenticity and the health of democratic processes.

If you are looking to craft a modern and digitally-driven political communication strategy, the GRA team can help: [email protected].