United States: generative AI in the 2026 midterm
The 2026 midterm is fought with growing use of generative-AI tools in campaign content creation and optimization, in an environment with no federal transparency law and debate over Section 230.
Documented electoral disinformation 2026
Tracking of electoral disinformation campaigns documented by organizations with open methodology (observatories, fact-checkers, observation missions) during the 2026 elections. It records the technique used (deepfakes, generative AI, coordinated networks, recycled content), the platform, the targeted actor and the source that documented it. It measures the gap between an apparently clean public discourse and the manipulation operations actually detected. It only gathers cases with methodological documentation, not suspicions; each record distinguishes between the campaign's detection and its attribution to an actor.
5
documented campaigns
Evolution
Data analysis
Statistical readings derived from the attributes of each recorded case. All figures come from the documented events; amounts are computed only over cases with a sum expressed in the indicated currency, without converting between currencies.
Disinformation technique
Platform
Region
Global incidence map
Choropleth by number of forensically or judicially documented cases. Countries with no verifiable public cases remain in the base colour — the absence of events does not equal the absence of surveillance. Hover or click a coloured country to see the cases.
The 2026 electoral disinformation no longer lives only in fake posts: it lives in AI-generated deepfakes, in coordinated networks that migrate to private messaging, and in ads reclassified to dodge transparency. This tracker gathers only what organizations with method have documented, always distinguishing between detecting a campaign and attributing it.
Disinformation is, by nature, hard to measure: its whole aim is not to be detected. That is why this tracker does not attempt to quantify 'all the disinformation' of an election —an impossible claim— but to record the campaigns and patterns that organizations with open methodology have managed to document: media observatories, digital forensic labs, electoral observation missions and fact-checkers. What is recorded is the tip of the iceberg that has been measured rigorously.
The 2026 map of techniques confirms a shift. Alongside the classic coordinated networks, deepfakes and AI-generated content now appear, plus a persistent migration toward private messaging —WhatsApp, Telegram— where moderation is almost impossible. In Hungary, the Digital Media Observatory warned of deepfake risk; in Colombia and Brazil, the weight falls on encrypted channels; in the United States, on generative AI applied to campaign optimization.
The tracker's golden rule: detecting is not attributing. Documenting that a deepfake or a coordinated network circulated is one thing; proving who financed and ordered it is quite another, and rarely possible with certainty. Each record separates the two, because conflating them turns a data point into an accusation.
Why it matters to those who watch
For a platform's integrity team, a fact-checker, a regulator or a researcher, this tracker's value lies in making techniques comparable across borders. When the same pattern —reclassifying ads as non-political, migrating to Telegram, generating AI avatars— appears in Hungary, Colombia and the United States in the same year, it stops being a national anecdote and becomes a global trend that can be anticipated.
That is the point of treating disinformation as data infrastructure rather than a succession of hoaxes: it lets you see the method behind the noise. And the method, unlike the specific content of each hoax, travels between elections and repeats. Documenting it with source and technique, election after election, builds the memory that platforms and regulators do not always keep.
Documented events (5)
Sweden: bracing for influence operations in 2026
Sweden is bracing for possible cyberattacks and foreign influence operations targeting the 2026 electoral process, amid regional tension and under the DSA framework.
Hungary: risk of deepfakes and illicit ads in the 2026 campaign
The Hungarian Digital Media Observatory (HDMO) warned of a real risk that the campaign would be influenced by a significant number of illicit political ads, including deepfakes, with huge reach, after the EU political-ad ban.
Colombia: private-messaging operations in the 2026 presidential race
WhatsApp groups and Telegram channels carried much of the campaign's narrative weight, with repertoires the Electoral Observation Mission has tracked in previous cycles. The 'Hondurasgate' broke into the conversation.
Brazil: the WhatsApp front in the 2026 general election
The WhatsApp + platform-liability intersection centers disinformation monitoring after the June 2025 STF decision that modified the liability regime, with the TSE watching coordinated networks.
Methodology
- event-log
- DC editorial construction
- event-driven
Each record documents an electoral disinformation campaign or pattern detected by an organization with open methodology during a 2026 election. The technique, platform, target and verifying source are noted. A distinction is drawn between the operation's detection and its attribution to a specific actor: many campaigns are documented without being attributable with certainty. Only cases with public methodological documentation are gathered; suspicions or unsupported accusations are not recorded. A country's absence does not mean the absence of disinformation, but the absence of methodical public documentation.