AI and Political Manipulation: Can the Right Not to Be Profiled Save Elections?
Dr. Luiz Silva, PhD's thesis proposes a new right to shield democracy from algorithmic microtargeting.
Political campaigns leverage artificial intelligence to exploit electoral vulnerabilities through microtargeting. Current laws fall short of providing protection, but Dr. Luiz Silva's thesis proposes a "right not to be profiled" as a solution. Drawing inspiration from environmental law, it aims to safeguard democracy's cognitive ecosystem.
In this article, we explore how contemporary political campaigns have abandoned the broad messages of the past to adopt microtargeting with absurd precision, utilizing artificial intelligence (AI) to exploit voter vulnerabilities. Our analysis is based on rigorous academic research: the doctoral thesis in Law, defended in 2025 at King's College London, by Dr. Luiz Silva, PhD. The study addresses deceptive influence as harm to democratic politics, proposing innovative solutions to protect electoral integrity.
From Megaphone to AI's Invisible Whispers
The era of broadcasting, where a single discourse was transmitted to all, is disintegrating. Previously, campaigns were like a "public theater," with clear messages open to collective debate. Today, we've moved from the "megaphone" to the era of microtargeted and hyper-segmented communications, where millions of "invisible whispers" are tailor-made for each person. AI transforms a campaign from a "giant billboard" into a "door-to-door salesperson," capable of changing its approach and convictions depending on the voter's hidden fears.
Dr. Luiz Silva, PhD's research reveals that AI can infer personality traits and psychological vulnerabilities from a vast "ocean of behavioral data" that seems harmless. This includes like patterns, screen time on certain images, vocabulary in casual posts, and even scrolling speed on a cell phone. By cross-referencing these micro-actions with established psychological models, AI deduces detailed profiles, even without identifiable personal data. With the help of generative AI, instant and personalized ads are created, exploiting fears (like crime or financial instability) or appealing to empathy and community, surgically shaping the perceived reality for each individual.
Why Current Laws Fail to Protect Democracy
Current laws, such as LGPD in Brazil and GDPR in Europe, were designed with a market-oriented mindset, primarily treating citizens as consumers. They aim to prevent the sale of data for commercial purposes but clash with freedom of expression when it comes to politics. Courts, fearing censorship of political debate, ultimately allow microtargeting. Furthermore, these laws are obsessed with identifiable personal data (name, CPF), failing to recognize anonymized behavioral patterns as vectors of manipulation, allowing AI to "fly through legal loopholes."
The Electoral Law is described as a "regulatory fossil," a "dinosaur." Built for the analog world, it is obsessed with tracking money spent on printed materials or television time, ignoring the near-zero marginal costs and massive reach of generative AI. The law is "looking in the wrong place," trying to count pennies while the threat operates at the level of psychological engineering and code manipulation.
The Right Not to Be Profiled: A New Shield for the Mind
Dr. Luiz Silva, PhD's thesis proposes an original contribution: the recognition of the "right not to be profiled." This new right is not based on market laws but on classic human rights, such as the right to respect for private life and, crucially, the right to freedom of thought and the voter's own freedom of expression. The idea is that if AI corrupts a person's original thoughts before they can even express them, freedom of thought is being invaded.
The proposal places this new right exclusively within electoral law, as it is the branch of law explicitly designed to balance the freedom of political communication with the protection of elections. Applying the necessity and proportionality test, the thesis demands an absolute and non-negotiable prohibition of profiling psychological and personality traits for electoral purposes. This prohibition is not censorship of content (politicians remain free to express their ideas) but rather of the predatory distribution method. For less intrusive demographic data (age, profession), an "opt-in" regime is proposed, allowing targeting with clear notice and explicit voter consent, preserving autonomy while shielding the psyche from manipulation.
Democracy by Design: Learning from Environmental Law
A central concept of the thesis is "democracy by design," which argues that protection should not be an afterthought but the foundation of any technological campaign system. For this, Dr. Luiz Silva, PhD, draws inspiration from environmental law, borrowing the precautionary principle. Just as a factory must prove it will not contaminate the ecosystem before operating, the thesis suggests that AI-driven disinformation is the "toxic waste" dumped into our "cognitive ecosystem."
This analogy fits perfectly: algorithmic manipulation degrades a shared informational environment, vital for our collective capacity for self-governance. The author argues that democracy cannot afford to wait for damage to occur before attempting to fine a campaign afterward. It is imperative to ban the "installation of the toxin duct" at the root, even before the electoral race, preventing the contamination of the shared reality that sustains democratic debate.
Final Reflection
Dr. Luiz Silva, PhD's thesis leads us from a bleak scenario of vulnerability to AI manipulation to a robust defense proposal. It maps the shortcomings of our laws and proposes an innovative right not to be profiled, based on human rights and focused on preventing harm to democracy's cognitive ecosystem. However, a challenging question arises: what happens if large segments of the electorate, aware of the risks, voluntarily decide to provide massive consent for microtargeting, preferring the "comfortable illusion of an algorithm" to the "arduous complexity of real debate"? Would the legal framework have the power to protect elections against the will of the voters themselves? This provocation raises a fascinating blind spot, where legal architecture confronts human nature, inviting continuous reflection on the invisible forces shaping the future of our society.