In the heart of Maputo, a quiet but profound recognition is settling in: the very tools designed to connect and inform us are now being weaponized to confuse and divide. Mozambique’s Prime Minister, Benvinda Levi, recently stood before regional leaders not to herald the dawn of a new digital age, but to sound a sobering alarm about its darker undercurrents. At a meeting of the Electoral Commissions Forum of SADC Countries, she laid bare the stark reality that the integrity of democratic processes across Southern Africa is under siege not by armies, but by ones and zeros. The proliferation of “deepfakes”—hyper-realistic but fabricated videos and audio—alongside automated disinformation campaigns poses a direct and existential threat to the public’s ability to discern truth from fiction, and by extension, to trust the systems that govern them. This is not a distant, hypothetical concern for a dystopian novel; it is the current, pressing condition of electoral governance.
Levi’s message was not one of pure doom, but of stark clarity. She framed the challenge as a defining test for the region’s electoral management bodies (EMBs), demanding they evolve from mere administrative entities into resilient, technologically adept guardians of public truth. The old model of ensuring secure ballot boxes and accurate counting is no longer sufficient. An election can be impeccably organized on paper, yet fundamentally compromised in the public square by a well-timed viral falsehood that labels a legitimate candidate as corrupt or claims a polling station is closing early. These phenomena, which she described as the “rapid technological change” and “digital manipulation” facing our democracies, require a new kind of institutional muscle. They demand that EMBs become “increasingly prepared, professional and resilient,” capable not just of running an election, but of fortifying the very information ecosystem that sustains it, all while navigating citizens’ growing demands for “greater transparency, inclusion and accountability.”
The Prime Minister’s address carefully articulated the dualistic nature of the technological moment. On one hand, she illuminated the transformative potential of Artificial Intelligence as a powerful ally in the pursuit of administrative perfection. She painted a picture of AI as a silent, efficient workhorse behind the scenes of democracy: revolutionizing the management of vast electoral registers, minimizing errors in voter rolls, and predicting logistical bottlenecks before they become crises. She spoke of its capacity to improve communication with voters on an unprecedented scale, delivering targeted, accurate information about registration deadlines or polling logistics directly to their phones. Furthermore, AI’s relentless power in data analysis could “facilitate the timely identification of operational irregularities,” spotting anomalous patterns in real-time data streams—a sudden, inexplicable surge in registrations in one district, for instance—that might otherwise signal fraud or systemic issues until it was too late. In this vision, AI is the ultimate administrative enhancer, strengthening the machinery of democracy from within.
But this optimism was earned through a clear-eyed acknowledgement of the flip side of the coin. With the same breath that she extolled AI’s virtues, Levi warned with stark precision: “the same technologies that can support democracy can also be used to weaken it.” She enumerated the specific, insidious tools of this digital subversion with forensic detail. There is the “abusive political micro-targeting” that uses stolen or misused personal data to deliver insidious, personalized messages designed to manipulate individual voters based on their psychological profiles. There is the relentless cacophony of “automated disinformation campaigns,” where bot armies drown out authentic voices and manufacture a false sense of consensus or outrage. And, most concerning, is the proliferation of deepfakes, which erode the very basis of credibility—if a video of a candidate saying something heinous can be faked, then real videos can be dismissed as fakes, plunging the electorate into a cynical reality where nothing can be believed. Levi argued for the “careful and responsible” use of these tools, insisting that they must respect legal mandates and, above all, safeguard citizens’ fundamental rights.
This tension brings the debate to the crucial, often overlooked battleground of data governance. Levi asserted that the ability to “collect, protect, manage and use” data safely and responsibly is not merely a technical issue, but a central component of a state’s “digital sovereignty.” This framing elevates personal data protection from a privacy preference to a matter of national integrity. In the electoral context, this becomes profoundly concrete: the integrity of the electoral register itself, the protection of voters’ biometric data (fingerprints, facial scans), and the security of computer systems are not just administrative details—they are the foundational pillars of democratic integrity. If a malicious actor can breach the system and alter registration records or steal personal data, the entire legitimacy of the subsequent vote is compromised. The Prime Minister’s message is clear: technological innovation cannot be pursued in a vacuum. It must be woven together with rigorous safeguards for privacy, system security, and democratic principles. Innovation without these guardrails is not progress; it is a vulnerability.
Acknowledging that these digital threats know no borders, Levi pivoted to a call for deeper, more synchronized regional solidarity. The challenge of a coordinated deepfake attack against a national election cannot be effectively countered by a single nation working in isolation; its impact can quickly ripple across a region, discrediting an entire regional bloc’s electoral credibility. She called for stronger cooperation mechanisms within SADC to “develop and strengthen regulatory and institutional frameworks” capable of ensuring the ethical, transparent, and responsible use of AI. This is a move toward a collective defense, where member states share intelligence on disinformation campaigns, collaborate on common standards for data protection, and build a regional pool of expertise to fortify each other’s electoral institutions. This regional front is essential, creating a united barrier against hostile forces that seek to exploit any weak link in the chain.
Ultimately, Prime Minister Levi’s speech was a profound meditation on the very essence of trust. She stressed that the credibility of an election does not hinge solely on its execution, but “above all, on the confidence that the process inspires in citizens, contestants and society.” This confidence is the most precious, and now most fragile, currency of any democracy. When public trust erodes, the legitimacy of the entire political system is thrown into question, paving the way for instability, conflict, and authoritarianism. By insisting that electoral bodies act with “impartiality, transparency, professionalism and respect for the law,” she reaffirmed that these timeless principles are the only reliable antidote to the chaos of this new digital age. Her closing remarks renewed Mozambique’s commitment to these democratic principles, but the call was to the entire region, a challenge to meet the future with a new kind of resilience—digitally fluent, deeply ethical, and fiercely protective of the public’s right to a free, fair, and truthful democratic process.