In the digital architecture of TikTok, fear does not simply pass through — it accumulates, compounds, and eventually converts. Justin Poncet, the architect of the European Union's anti-fake news algorithm, describes a field where social media platforms are no longer conduits for communication so much as arenas where the contest for truth is fought, and where the outcome is rarely neutral.

"TikTok has a unique ability to amplify emotions," Poncet said in an interview. "Its algorithm is designed to keep users engaged by promoting content that elicits strong reactions, and fear is one of the most powerful emotions it can exploit." What he is describing is not a technological accident or an unforeseen side effect. It is, in his framing, a calculated feature — one with consequences that extend well beyond the screen and into the polling booth.

The mechanism is worth dwelling on. Engagement-optimized algorithms do not distinguish between a reaction that informs and one that merely agitates. They measure duration, shares, comments — the residue of emotional activation. Fear, which is faster to trigger and slower to dissipate than most other responses, performs exceptionally well by these metrics. A platform built to maximize engagement is, almost by structural necessity, a platform that rewards the most alarming version of any story.

Romania enters this picture not as an outlier but as a clarifying case. Social networks are the primary source of information for a substantial portion of the Romanian population — millions of people whose understanding of political reality is shaped more by what surfaces in a feed than by what appears in a newspaper or on a broadcast. "The spread of disinformation here is not just a matter of misinformation," Poncet explained, "but a structured attempt to sway public opinion and, ultimately, votes." The distinction matters: misinformation can be accidental; what Poncet is describing is architecture.

Forensic doctors in Romania initiate warning strike

US Apache Shot Down Amid Rising Tensions in Strait of Hormuz

The conditions that make Romania particularly susceptible are not difficult to trace. A history of political instability, a population with deep and historically grounded skepticism toward traditional media, and an information environment in which social networks have become the de facto news outlet for millions — these are not incidental features. They are the substrate on which disinformation campaigns are cultivated. "Disinformation campaigns are often tailored to local contexts, exploiting existing fears and biases," Poncet noted. "In Romania, these campaigns can be particularly potent given the societal backdrop."

The consequences, as Poncet elaborated, are not confined to individual beliefs. Disinformation erodes trust in democratic institutions, distorts electoral outcomes, and widens the fractures that already run through a fragile political environment. "It's not just about misleading voters," he said. "It's about undermining the very fabric of democracy." That is a large claim, but it is one grounded in observable patterns: when the information environment becomes sufficiently polluted, the capacity for collective judgment — the precondition of democratic legitimacy — begins to degrade.

Poncet's algorithm represents one response to this problem, a tool designed to identify and flag fake news at scale. But he is careful not to overstate what technology can accomplish. "Education is key," he said. "We need to equip people with the skills to critically evaluate the information they encounter online." The algorithm can surface a pattern; it cannot, by itself, change the habits of mind that make a population vulnerable to that pattern in the first place.

What Romania's experience clarifies — and why it is, as Poncet suggests, a relevant case for understanding disinformation across Eastern Europe — is that the problem is neither purely technical nor purely political. It lives at the intersection: in the gap between the speed at which false information travels and the much slower speed at which institutions, educators, and citizens can respond. Social networks became the main channel for information consumption for millions of people faster than any regulatory or civic framework could adapt. That gap is where fear finds its use, and where votes are quietly, structurally shaped.

The question that Romania's case leaves open is not whether disinformation can be countered — Poncet's work, and the broader field of platform accountability research, suggests it can. The question is whether the response can move at the same speed as the problem it is trying to solve.