What should be the role of personal data in campaigning?

By Joshua Vickers

Britain’s parties have perfected data-driven campaigning. What they haven’t mastered is accountability. 

In a terraced house in Leeds, a pensioner opens Facebook to check on the grandkids and is confronted by a blue-and-white ad warning of crime surges and benefit fraud. Three hours down the M1, in Hackney, a student scrolling TikTok encounters a slick, meme-laced video about climate justice tagged #VoteForChange . Neither realises they were selected—not by chance, but by an algorithm. 

British elections are no longer fought on doorsteps. They unfold in newsfeeds and inboxes—via pixels and profiles. What began in the 2000s as simple voter databases— Labour’s Contact Creator, the Conservatives’ Merlin—has become a data arms race. Parties now hold millions of records on voters, segmenting the electorate with the precision of Silicon Valley marketers and the subtlety of spies. 

The breakthrough came in 2015. The Conservatives spent over £1.2 million on Facebook ads—ten times more than Labour. They borrowed tactics from Obama’s US campaign playbook, combining postcode data with emotional micro-messaging. Swing voters received “synthetic intimacy,” not broad manifestos. Labour knocked on doors. The Tories knocked through screens. 

A Clean Break—With the Rules. 

Then came Brexit. Vote Leave’s £1.5 million partnership with AggregateIQ delivered over 1,000 distinct Facebook ads, micro-targeted around fears of immigration and NHS collapse. Some warned of Turkey’s imminent EU entry. Others pushed the infamous £350m NHS lie. Behind it all, a covert pipeline: Vote Leave funnelled funds through BeLeave, its youth affiliate, breaching campaign finance laws. The Electoral Commission fined them and referred the case to police. “Win at all costs,” said one insider—it was more strategy than slogan.

By 2019, data-driven campaigning had metastasised. Major parties spent over £9 million on digital ads. The Tories flooded “Red Wall” constituencies with repeated blasts of “Get Brexit Done.” Labour, with Momentum, carpeted Instagram and Facebook with NHS tributes and tuition promises. Third-party “outrider” groups spent an additional £2.7 million—largely untraceable. 

Profiling grew darker. The Conservatives used credit data and even name-based ethnic estimations from Experian to classify 10 million voters by race and religion. The Information Commissioner ruled this illegal. The penalty? None. “A significant shortfall in transparency,” the ICO said. Critics called it cowardice. 

Tools like NationBuilder—once dubbed the ‘Obama startup’—gave even local candidates Silicon Valley-grade reach. Political parties now target “Worried Young Renters” or “Affluent Achievers” based on Mosaic codes—homeownership, credit bands, even online behaviour. Campaigns utilise Facebook Pixels and cookies to build digital dossiers. AI models infer who might be persuadable—and how.

Transparency, meanwhile, is elusive. Facebook’s Ad Library displays sponsors, not targeting logic. Google provides ad archives, but their detail is laughable. TikTok prohibits political ads, yet both Labour and the Tories promote memes through influencers. It’s targeting by stealth. What voters see—and what others don’t—is invisible by design. 

The Regulators Blinked. 

The law exists, nominally. The 2022 Elections Act introduced “digital imprints.” UK GDPR demands a legal basis for data profiling. However, the watchdogs rarely bite. The ICO uncovered unlawful profiling, yet fined no party. Former commissioner Elizabeth Denham told Parliament, “We are not here to punish.” The parties, it seems, took note. 

Enforcement, where it exists, is performative. In 2019, the Conservative press office rebranded its Twitter account as “factcheckUK” mid-debate, pushing partisan talking points under the guise of neutrality. They cloned Labour’s manifesto site to deceive Google searchers. Both stunts were condemned, neither punished. As one MP quipped, “The law shrugged.” 

The next disruptor is not a party—but a processor. In Brighton, a candidate named Steve Endacott ran with an AI persona, “AI Steve”, fielding real-time queries from thousands of voters. Deepfakes of Putin addressing UK voters in English have circulated. The line between persuasion and manipulation is vanishing. Soon, the candidate knocking may be synthetic—and the conversation pre-determined. 

Platforms comply just enough to escape scrutiny. Facebook and Google offer ad libraries—vague and largely outdated. YouTube has promoted conspiracies, while TikTok boosts any content that gets views, irrespective of accuracy. A 2023 test by Global Witness submitted fake UK election ads to all major platforms, with only TikTok and YouTube rejecting them. In India, those same platforms approved everything.

Power by Proxy. 

Post-Brexit, Britain is at a crossroads. The EU is prohibiting micro-targeting driven by sensitive data. In contrast, Britain’s Data Protection and Digital Information Bill may loosen protections, making it harder for citizens to understand what information parties hold about them. 

British democracy is not yet broken. But it is brittle. “Systemic vulnerabilities,” the Electoral Commission warned in 2020, remain unpatched. Voters are profiled, segmented, and served different realities. If data is the new battleground, then accountability is AWOL. 

Data can enlighten, energise and enfranchise. It can also distort, divide and deceive. Campaigns will grow more digital, and voters more finely profiled. The next election will not be won in Parliament, nor lost on the doorstep. It will be processed in servers, ranked by engagement, and served through opaque feeds. Voters will not share the same promises, nor the same fears—only the illusion of choice. 

If democracy is to be more than performance, its rules must be rewritten before it becomes just another dataset. 

Image: Synthetic Intimacy, Joshua Vickers