For most of the past three decades, U.S. domestic-terrorism conversations have been framed by two patterns: the long, deadly arc of far-right violence and the episodic but devastating mass-casualty plots associated with jihadist inspiration. In 2025, that familiar mental map hit an unexpected jolt. A new, data-driven assessment from researchers at the Center for Strategic & International Studies (CSIS) indicates that left-wing terrorism incidents have climbed sharply—enough that, for the first time in more than 30 years, left-wing attacks and plots are outnumbering those from the violent far right. This reversal is not a declaration of “new normal” so much as a flashing indicator on the dashboard: the underlying engine of American extremism is responsive to political context, grievance cycles, and the oxygen we give to narratives of existential threat.
The claim deserves careful handling. “Left-wing terrorism is rising” can be instantly weaponized in partisan discourse, either to minimize years of far-right lethality or to score a rhetorical win in a news cycle shaped by high-profile political violence. CSIS’s researchers, however, place the surge in context. Their longitudinal dataset—about 750 attacks and plots from 1994 through mid-2025—shows that left-wing incidents are indeed elevated this year, but from historically low levels. In absolute terms, far-right violence over the past decade remains far more lethal. The point, then, is not to flip a hierarchy of danger but to recognize that opportunity structures for political violence shift, and that the U.S. counterterrorism posture must be nimble enough to track those shifts without losing perspective.
A Reversal With History
The CSIS finding matters because it jars our assumptions. Since the Oklahoma City bombing and through waves of militia resurgence, anti-government, white supremacist, and accelerationist currents have accounted for the preponderance of deadly violence on U.S. soil. Even after 9/11 redirected national security toward jihadist threats, domestic-terror fatalities in the 2010s and early 2020s disproportionately stemmed from far-right actors. In the first half of 2025, though, CSIS charts a spike in far-left incidents tied to sabotage, arson, property destruction, and targeted, smaller-scale assaults—tactics that prioritize disruption and symbolism over mass casualties. The pattern echoes earlier U.S. eras (e.g., 1970s political violence) in methods, even if the modern ideological cocktail differs.
Several contextual notes keep this from being misread. First, incident counts (attacks and plots) are not the same as lethality. A year with numerous non-fatal sabotage attempts can eclipse a period with fewer but deadlier shootings. Second, “left-wing” is a heterogeneous bucket: eco-sabotage, anti-fascist vigilantism, anarchist direct action, and anti-police militancy often mingle tactics while diverging on ultimate aims. Third, CSIS is explicit that the reversal could be temporary: a contraction in far-right plots during 2025 could snap back, particularly in a charged election environment, as grievance entrepreneurs pivot strategies. The picture is dynamic, not deterministic.
Political Context and Narrative Fuel
Why now? Analysts point to a confluence of narrative, policy, and policing dynamics. The White House’s posture toward political violence, the rhetorical elevation of certain threats, and episodic shocks (including a prominent conservative political figure’s assassination, which supercharged media frames) all reconfigure the incentive landscape for fringe actors. The Trump administration’s declaration that “antifa” constitutes a domestic terrorist organization—despite the term’s nebulous, non-hierarchical reality—also altered how incidents are labeled, policed, and reported. That combination can both chill and provoke: it deters some actors while galvanizing others who interpret the designation as proof of encroaching authoritarianism.
Meanwhile, the years-long discourse about far-right violence has had twin effects. On one hand, it drove serious investments into monitoring and interdicting plots linked to white supremacists, militias, and accelerationists, potentially depressing incident counts in early 2025. On the other hand, it fostered a mirror-narrative among radical left currents that “the state protects fascism,” a justification used to escalate from protest and civil disobedience to direct action targeting infrastructure, political offices, or perceived collaborators. The result is not moral equivalence but a reminder that counterterrorism is a system of interacting narratives: the story each side tells about the other can function as recruitment material.
Method Mix: Sabotage Over Slaughter
Methodology differentiates the current far-left spike from prior far-right surges. Where far-right violence in recent years often sought body counts (e.g., mass shootings targeting synagogues, Black churches, immigrant communities), the contemporary far-left profile skews toward property destruction, arson, and intimidation—tactics that create fear and signal capacity while stopping short of indiscriminate killing. In CSIS’s coding, these are still terrorism if they aim to coerce broader populations or governments for political ends. But the risk curve is different: a campaign of infrastructure sabotage produces cumulative societal costs and psychological pressure without necessarily triggering the acute, front-page horror of a mass-casualty event. This complicates media coverage, policing priorities, and public risk perception alike.
What follows from that method mix is a policy conundrum. Traditional counterterrorism has grown adept at thwarting catastrophic plots—especially those leaving digital trails, supply-chain footprints, or large conspiratorial chatter. Dispersed sabotage by loosely affiliated actors is harder to detect because it demands little coordination, requires modest resources, and can be framed (internally) as morally bounded: “We hit property, not people.” The difficulty is that escalation pathways exist. “Non-lethal” tactics can bleed into assaults, boobytrapping, or reckless arson with unpredictable human consequences. The imperative for law enforcement, then, is precision: aggressive enough to deter escalation; restrained enough to avoid broad-brush crackdowns that radicalize sympathizers.
The Designation Trap
One news-driving feature of 2025 has been the political impulse to “designate” domestic adversaries. Labeling “antifa” as a terrorist organization may satisfy a demand signal from parts of the electorate, but it raises practical and civil-libertarian concerns. Unlike foreign terrorist organizations, domestic movements lack a centralized legal target for proscription; the First Amendment constrains prior restraint, association penalties, and the criminalization of ideology. The effort to apply designation logic domestically risks two outcomes: (1) it muddies analytic clarity by collapsing diverse actors under a catchall brand; and (2) it invites reciprocal escalation when a future administration uses the same tool against an opposing fringe. Given that CSIS’s data suggests a malleable threat landscape, building durable, content-neutral guardrails around political-violence suppression may be wiser than swinging the designation hammer.
The United Kingdom provides a cautionary parallel. In 2025 the government proscribed Palestine Action under terrorism law, and police subsequently made arrests for “supporting a proscribed organization” at protest actions—sparking intense debate over the line between criminal conspiracy and protected dissent. The British context is not the U.S. Constitution, but the episode underscores how terror law can quickly extend from violent acts to expressive support, chilling speech and inflaming grievances. It also demonstrates the enforcement reality: once an organization is proscribed, otherwise-lawful acts (e.g., advocacy, fundraising, even certain kinds of praise) can trigger serious penalties—outcomes that can be both effective against violent facilitation and counterproductive for broader social peace.
Media, Measurement, and the “Which Violence Counts?” Problem
A second trap is analytic: in polarized environments, every dataset becomes a cudgel. One camp cites the CSIS finding to argue that left-wing extremism is finally being “seen”; the other points to a longer series of studies showing far-right violence as more frequent and more deadly over time. Both can be true—and, in fact, both probably are. Datasets differ in scope (plots vs. attacks), definition (what counts as terrorism vs. hate crime vs. riot), and observability (what’s reported, charged, or covered). Even within a single dataset, a shift from 10 to 30 incidents can look dramatic in percentage terms while remaining low in absolute risk compared to a previous year’s 5 mass-casualty attacks by the opposition. Good analysis foregrounds those caveats and resists cherry-picking.
The media dynamics are equally fraught. Partisan outlets highlight different slices of the same research, sometimes tying them to specific, emotionally resonant events. That can help mobilize attention and resources, but it can also distort public understanding of baseline risk. The proper response is not symmetrical coverage for its own sake, but a clear hierarchy of threats based on lethality, demonstrated capability, target selection, and escalation potential—updated in real time as those variables move.
What Drives the Far-Left Spike?
Three drivers recur in expert commentary:
-
Catalytic Grievances. Policing controversies, environmental flashpoints, abortion politics, and perceived authoritarian drift can accelerate recruitment into direct-action networks. These grievances are often episodic but can stack into a durable identity.
-
Network Tactics. The lack of formal hierarchy (cells, affinity groups, online collectives) reduces vulnerability to infiltration and decapitation. It also increases variance in discipline and risk tolerance.
-
Policing and Prosecutorial Adaptation. As law enforcement devotes significant bandwidth to far-right and jihadist threats, a portion of far-left actors perceive a permissive environment for low-grade sabotage—until enforcement catches up. When it does, heavy charges can backfire if the public perceives disproportion.
These are not moral defenses; they are operational factors. Understanding them is necessary to design interventions that cut off escalation pathways without turning run-of-the-mill protest into a pipeline to radicalization through overbroad surveillance or indiscriminate crackdowns.
Policy Implications: A Balanced, Flexible Posture
If left-wing incidents are up in 2025, what should policymakers and practitioners do—especially without losing sight of far-right lethality?
First, keep threat accounting honest. Policymakers should track multiple metrics—incident counts, plots disrupted, fatalities and injuries, target categories, weapon types, and ties to organized networks. Present that mosaic publicly, with methodological notes. Reaffirm that prioritization is about demonstrated harm and realistic risk, not partisan identity.
Second, reinforce content-neutral legal tools. Criminalize conduct, not ideology. Material-support statutes, conspiracy law, arson, sabotage, weapons offenses, and RICO-like tools can address violent facilitation across the spectrum. Reserve “designation” rhetoric for foreign organizations, where Congress has already created a constitutional framework. Domestic speech and association—no matter how odious—should trigger criminal exposure only when linked to concrete, unlawful acts.
Third, target escalation chokepoints. Focus on behaviors that convert protest into terrorism: procurement of incendiary materials; reconnaissance of critical infrastructure; cross-state conspiracy; online tutorials that shift from “resistance” to explosive construction. Narrow, high-confidence enforcement prevents both overreach and under-reach.
Fourth, invest in preventive ecosystems. Community-level violence interruption, credible-messenger programs, and exit pathways are usually discussed for gang violence and jihadist deradicalization; they have analogues for political violence too. Create grants that are ideology-agnostic but behavior-specific—offered to groups that can de-glamorize sabotage and highlight non-violent, high-impact civic pathways.
Fifth, align media and public-information strategies. Law enforcement and civic leaders should avoid sensationalizing low-yield sabotage while also refusing to euphemize violence. Transparent, specific briefings build credibility: “Here’s what happened. Here’s why it’s terrorism under the statute. Here’s what we’re doing next.” Precision language blunts propaganda.
Sixth, recalibrate infrastructure protection. If far-left actors are targeting critical equipment (construction sites, energy nodes, offices of political parties), threat modeling should prioritize soft-target hardening and rapid repair capacity. For private firms, information-sharing with state fusion centers and DHS becomes crucial—again, on a content-neutral basis that protects lawful dissent.
The International Mirror
The U.K. case around Palestine Action illustrates how liberal democracies wrestle with “where protest ends and terrorism begins.” British authorities banned the group as a terrorist organization; subsequent arrests included individuals accused of “support” for a proscribed entity. Advocates worry about chilling effects, while the state emphasizes the seriousness of sabotage against military-linked targets. The U.S. legal architecture differs, but the strategic dilemma rhymes: aggressive criminalization can suppress violence in the short term yet deepen grievance narratives that seed the next wave. The lesson is to maintain tight legal definitions and evidentiary discipline, especially when the theater of conflict includes symbolic acts calibrated for media impact.
Holding Two Ideas at Once
Two propositions can live together without contradiction. First, far-right violence has been—and may again be—the most lethal and strategically destabilizing U.S. domestic-terror threat of the modern era; that must remain central to policy. Second, the 2025 rise in left-wing terrorism incidents is real enough, and norm-shifting enough, that dismissing it as “statistical noise” is a mistake. The security state must be capable of walking and chewing gum: deterring emergent left-wing sabotage while never taking eyes off far-right networks with a proven appetite for mass murder.
A final note: Political violence is downstream of political culture. When leaders normalize apocalyptic narratives, flirt with eliminationist rhetoric, or theatrically designate amorphous internal enemies, fringe actors on all sides feel licensed to escalate. Conversely, when leaders communicate that democratic change is still available—and that rivals are opponents, not enemies—the oxygen available to violent entrepreneurs diminishes. The CSIS dataset is not a prophecy; it is a temperature check. It tells us we still have agency over the climate.
References (APA)
Axios. (2025, September 28). Study: Left-wing terrorism climbs to 30-year high. https://www.axios.com/2025/09/28/left-wing-terrorism-far-right-violence-research (Axios)
Byman, D., & CSIS Warfare, Irregular Threats, and Terrorism Program. (2025, September). Left-Wing Terrorism and Political Violence in the United States: What the Data Tells Us. Center for Strategic & International Studies. https://www.csis.org/analysis/left-wing-terrorism-and-political-violence-united-states-what-data-tells-us (CSIS)
CSIS Warfare, Irregular Threats, and Terrorism Program. (2025, September). Ideological Trends in U.S. Terrorism. Center for Strategic & International Studies. https://www.csis.org/analysis/ideological-trends-us-terrorism (CSIS)
Fadel, L., & Tong, S. (Hosts). (2025, September 25). Study finds left-wing political violence on the rise [Radio broadcast]. WBUR/Here & Now; TPR syndication. https://www.tpr.org/2025-09-25/study-finds-left-wing-political-violence-on-the-rise (TPR)
Parker, A. (2025, September 25). Left-wing actors responsible for more attacks this year, new research indicates. The Washington Post. https://www.washingtonpost.com/nation/2025/09/25/political-violence-leftist-right-wing/ (The Washington Post)
The Atlantic (Editors). (2025, September). Left-Wing Terrorism Is on the Rise. The Atlantic. https://www.theatlantic.com/politics/archive/2025/09/charlie-kirk-left-wing-terrorism/684323/ (The Atlantic)
Williams, V. (2025, September 20). Analysis: What data shows about political extremist violence—far-right attacks remain more deadly. PBS NewsHour. https://www.pbs.org/newshour/politics/right-wing-extremist-violence-is-more-frequent-and-deadly-than-left-wing-violence-data-shows (PBS)
Townsend, M. (2025, September 28). Dozens arrested at Palestine Action protest outside Labour party conference. The Guardian. https://www.theguardian.com/uk-news/2025/sep/28/arrests-palestine-action-protest-labour-party-conference (The Guardian)
Baehr, J. (2025, September 26). Charlie Kirk killing puts rise in left-wing terror in spotlight as study shows violence hitting 30-year high. Fox News. https://www.foxnews.com/us/charlie-kirk-killing-puts-rise-left-wing-terror-spotlight-study-shows-violence-hitting-30-year-high (Fox News)
(Note: The body of the essay contains no hyperlinks; sources are listed here for verification.)
Posts
