On the off-hand chance you’ve just suffered a severe concussion, 2025 was a year of immense uncertainty. Even before Donald Trump’s inauguration, the start of the year saw a firebombing with a Tesla Cybertruck in front of the Las Vegas Trump Hotel. Matthew Livelsberger’s motivations appear to have been unique to himself, rooted in personal narratives likely connected to his struggles with PTSD. The event marked a theme that would continue through the year, highly personal narratives leading to individual actions difficult to connect to larger ideologies, but nonetheless driving the temperature in contemporary American society higher and expanding the possibility space for violence.

2025 cemented a pattern that arguably began in earnest in 2024 with two attempted assassinations targeting Donald Trump and the assassination of UnitedHealthcare CEO Brian Thompson late in the year. It’s contrasted with high-profile and obviously partisan protests across the nation which saw virtually no violence connected to them, even in the context of more kinetic protests against ICE.

Looking to 2026, we’re left with two questions. How do we square the circle of this apparent contrast and what does this mean for the potential for increased risk in the coming year? Specifically, so this isn’t a small book, what does this say about the potential for violence and disruption of the 2026 midterm elections? The answer is, unfortunately, very simple. As violence becomes more connected to personal grievances and narratives created in increasingly small, isolated corners of the internet, it becomes less predictable but with a much lower barrier to entry and, thus, becomes considerably more likely.

In the specific terms of a threat outlook, for now, there is very little good news and what there is rests almost entirely on the word “relatively”.

The 2026 threat landscape is defined by a pair of categories divided by their motivation, those with clear political or ideological motivations and those with personal or non-ideological motivations.

The primary analytical distinction between these categories is that networked, ideological actors are easier to predict and less likely to initiate large-scale action without an outside force while individual, less-connected actors are much more unpredictable and are more likely to act without identifiable provocation. This combination of threats suggests a fringe prospect of individual, stochastic violence leading to larger-scale violence from more organized parties in response.

As we approach the start of primary season, concerns are already being raised about the prospect of election interference. Obviously, such interference as a “defense” against false narratives of Democrat-led election interference appears to fit well within the playbook of the current administration, but it is worth remembering 2025 saw the GOP suffer significant electoral defeats with no pushback.

Moreover, the publicly-understood reality already appears to be that the GOP will suffer significant losses in 2026. This lack of structure for a reality in which the GOP does well in 2026 and a surprisingly incompetent second Trump administration makes large-scale election interference perhaps not unlikely, but certainly less likely.

This is not to say it won’t happen. It is very likely that closely contested elections will be aggressively challenged in court or face extreme and hostile scrutiny, but pervasive election interference appears, at time of writing, relatively unlikely. That likelihood will shift significantly if elections suffer significant disruptions from outside sources.

As stated in a 2024 US intelligence report which noted the primary threat to the then-coming election was violence from individual actors motivated by conspiracy theories and other, more personal narratives rather than militias. This is likely to be the case in 2026. While militias like the Oath Keepers claim to be reforming in preparation for the election and the potential for political violence, these organizations primarily center themselves around hypothetical violence as a reaction to outside forces. In 2025, decentralized groups operating in cells, like active clubs, have proven a more significant threat.

Militia organization and early mobilization can be difficult to observe from the outside, but their networking is often apparent. Groups like The Old Glory Club form an inter-organizational support network between high-profile individuals in militia spaces and other prominent far-right actors.

Historically, organized militias do not present a significant, systemic danger without an existing milieu of widespread violence. For our purposes, that means if we enter into a scenario in which militias present a significant threat, it will come after a period of escalation that will be obvious as it happens. While one could argue this environment already exists, the lack of mobilization implies it is simply not severe enough. It is unlikely militias will be a primary source of violence. If stochastic, individual violence or large-scale, disruptive protests occur more frequently in 2026, the possibility of militia-driven violence will predictably increase.

In 2024 and through 2025, a series of shootings beginning with Thomas Crooks’s attempted assassination of Donald Trump increased the profile of a developing dynamic in self-radicalization. This pattern continued with Ryan Routh, Luigi Mangione, Vance Boelter, Tyler Robinson, and Joshua Jahn. A uniting characteristic of these shooters is that they do not match the “traditional” model of radicalization. Instead, these individuals lack any clear connection to an extremist network or clearly articulated political views that can be easily mapped onto a broader ideological framework.

These shooters appear primarily motivated by individual, personal narratives only marginally connected to outside political ideology, representing an evolving form of stochastic violence. Though there is not significant research into this motivation for violent action, it appears to be a product of degrading public rhetoric and an increasing acceptance of violence as a necessary end-product of that rhetoric. This narrative is potentially appealing to socially isolated individuals seeking an expression of a personal heroic narrative.

Given the frequency of public appearances by political candidates and large public gatherings with complex security demands and difficulty in observing lone actors, this threat is unpredictable, but its opportunistic nature makes it likely.

Throughout 2025, doxxing and bomb threats have increased in frequency and appear largely attributable to small online networks, some of which are connected to a larger milieu of antisocial and misanthropic violence. Not all of this activity is attributable to larger networks like Com as false bomb threats and doxxing have a low barrier to entry and can be achieved by virtually any unsophisticated actor.

In November of 2025, an election in New Jersey was targeted by a bomb threat allegedly committed by a local juvenile. This form of disruption requires minimal resources and expertise. Critically, while an individual may be encouraged by a network, this style of disruption does not require participation in a network.

The combination of ease of engagement and agnosticism towards network participation creates a highly dynamic, unpredictable threat with a low barrier to entry. It is very likely we will see bomb threats and doxxing connected to the 2026 primary and general elections.

Over the last two years, increasing attention has been paid to the 764 network as part of the larger Com network. Violence produced within these networks do not specifically target elections, instead targeting children for misanthropic violence, extortion, and grooming into antisocial violence. However, 764 and criminal acts which can be attributed to it (like the swatting group mentioned above) will likely figure into narratives around the election.

It is very likely high-profile violence committed as part of 764/Com will occur during 2026, but its exact nature and timing are inherently unpredictable. Notably, 764 and Com target at-risk and marginalized teens and young adults, making it likely narratives relating to these networks will connect to other narratives involving gay and transgender people, narratives which will play a central role during 2026.

Beyond 764, accelerationist groups continue their activities. High-value targets of accelerationist violence include critical infrastructure as part of a larger goal of the destruction of society on a fundamental level. The process and infrastructure of an election is just such a target and given the increased activity in accelerationist networks, an attack on that infrastructure appears probable. As these networks select targets for maximum harm, such an attack would be more likely to impact election processes.

Throughout 2025, significant protests occurred across the United States in response to abductions by ICE and the increasing unpopularity of Donald Trump’s policies. Most notable among these are the “No Kings” protests which so far have seen historic attendance and a near-total lack of violence. This has not prevented hostile and violent rhetoric targeting these events.

Protests are a soft target for violent actors, particularly vehicle ramming attacks given their large attendance and ease of access. It is certain there will be more large-scale protests during 2026 and, as narratives around the election grow more intense, so too does the probability of violence targeting these events.

Obviously, much more than voting will happen in 2026 and while this article represents an effort in constraint, it’s worth considering two major threat vectors that characterized 2025 and will likely continue to characterize 2026.

At time of writing, legitimate daycares in the United States are the target of conspiracies and violent hostility relating to an ongoing fraud investigation which has recently been co-opted for obvious political purposes.

This is a pattern we have seen before and that will likely continue in 2025. A real event—SBA fraud in this case—narrativized in spaces like TikTok, X, and podcasts into a full-blown conspiracy that must be aggressively countered. While this could be lumped in with earlier discussion of heroic narratives, I would make the case this practice is too public and too social to fit. Instead, this has more in common with QAnon and other social milieus which heavily rely on conspiracy theories and the perception of a large conflict against an ambiguous but ever-present threat.

A similar event recently played out in the aftermath of the Brown and MIT shootings with participants in social media engaging in widespread speculation and theorizing about coverups, incorrect identification of potential suspects, and general rampant narrativization. Again, this appears to be a product of contemporary social media and it is likely this will be a growing threat in 2026, especially as the motivations to create narratives around election fraud and voter suppression become more immediate.

There is truly nothing I can write that hasn’t already been written by a legion of pundits and tech journalists, so I’ll be brief.

There is no real utility to generative tools that is not inherently deceptive and, with the right motivations, immensely destructive to shared reality as a concept such as it existed before 2025. 2026 will see more deepfake scandals, more claims that real events are deepfaked, and the aggressive eroding of what remains of public trust and cohesion.

Worse, if the AI bubble pops in the coming year, the economic destruction wrought by the event will greatly increase societal friction with such deep and wide impacts as to be almost completely unpredictable. It is, in this way, a singularity—a point beyond which our models collapse and all prediction is pure speculation.

On the whole, the outlook is exceptionally grim, but I would ask you to keep in mind some of that is core to the nature of a threat landscape. This has been an exercise in looking for bad news. If you are online enough to be reading this in the first place, you know very well there is always bad news if one goes looking for it.

I could—but will not—spend three times the length of this article considering the potential social and political ramifications of these and other events and, to my mind, most of those outcomes are a net positive.

2026 will be another year of significant challenges and much greater uncertainties. While it is a tautological non-statement to say 2027 will be defined by what happens in 2026, it really is very difficult to say where we’ll be in January of next year. We can, at the very least, make some educated guesses about the shape of events that will define what we’ll be worrying about.

Between now and then, I’ll try to update this blog more than once.

No promises.