Expectations

Back in 2016, a D.C. think tank approached me to write a short future-fiction style vignette for a project that never got off the ground. It’s been sitting on my hard drive ever since, but last weekend’s horrific Hamas attack on Israel, the specter of still more civilian death to come as Israel retaliates, and the thus-far and yet-to-come spread of graphic content on social media got me thinking about it again.

The 2016 vignette was formatted as a faux memo from a future administration describing an extremism-related incident. I have included it below, very lightly edited. After, I want to talk a little bit about what it got right and what it got wrong.

____________________________________________________________________________________

Date: 9/12/2026

From: National Security Advisor for Social Network Security

Subject: Scarlet Friday

Domestic Identity Movements And Violent Unrest At Crisis Point

The event began around 11 a.m. Eastern time when the New Islamic State sent a mass broadcast of a graphic video showing the execution of American hostages. The video was sent directly to approximately 3 million cell phones across the United States, along with a threatening message. The attack was distributed, employing malware to send the message from approximately 10,000 infected phones to people on the sender’s contact list. Because the message appeared to originate with a friend or acquaintance, many recipients opened the video and viewed it, infecting their own phones with malware and furthering the transmission.

The broadcast was not randomly targeted. The malware was initially spread by the hack of an app known as the White Genocide Reporter, distributed through white nationalist forums, which provides news and links curated by white nationalist social media users. As a result, recipients of the message were primarily white nationalists and their friends and families. While Friday’s events were tragic, it is perhaps fortunate that only 0.03 percent of the video’s recipients responded violently. But because of the scale of the transmission, that amounted to 900 people who carried out some form of violent hate crime against Muslims and other immigrants after the video was sent. While the majority of these were small-scale actions that did not result in loss of life, a small percentage created disproportionate havoc, with 41 mass shootings and 13 bombings taking place in the course of a single day.

The chaos triggered alerts on the Actioner activist social network, which is used by a wide variety of groups to rapidly mobilize supporters. Thousands of activists with various right- and left-wing affiliations clogged highways and public transit traveling to areas most impacted by violence, often crossing state lines. While many demonstrations were peaceful, fighting broke out in several locations among radical groups representing a variety of ethnic and religious identities. In at least two cases, demonstrators exchanged gunfire with police at a significant scale.

By the end of the second day, nearly 4,000 people were dead in hundreds of separate violent incidents.

We assess that this attack was intended to spark a wider race war, similar to the intent of well-known incidents in 2017, 2019, 2023, 2024 and twice in 2025. Then, as now, American values of diversity and pluralism ultimately prevailed, despite weeks of elevated tensions after each incident. But the significant minority of Americans now meaningfully participating in exclusive ethnic and religious identity movements is approaching 9 million. When members of these groups act in a synchronized or cascading manner, the challenge to public safety is critical.

Thwarting Unrest: New Measures To Reinforce America’s Social Center

Without an organic corrective movement to strengthen the social center, additional outbreaks of violence are inevitable, and the threat in the short term is likely to be significant. It would have been better if this effort had started sooner. After the bitterly divisive elections of 2016 and 2020, it was clear that significant segments of American society had become fragmented and polarized along identity lines. These elements, while still small relative to the overall U.S. population, were able to leverage emerging social technologies to create ideological news sources (with significant assistance from Russia and other hostile state sponsors), at the same time that traditional mainstream news outlets contracted under severe pressure from the third wave of the disruptive economic transition to digital and online distribution. ?

There are limits to what the federal government can do to address this problem. Obviously, we can continue our efforts to counter the efforts of hackers and cybercriminals, in order to patch the vulnerabilities exploited in this attack. But the hack is ultimately a mechanical challenge. The underlying problem is the long-term trend toward fractionalization of the American body politic. Naturally, there will be calls to harden physical security in the wake of Friday’s events, but the greater vulnerability is social, both in the number of individuals who now gravitate toward identity-based extremist movements and their increasing tendency to mobilize in violent action.

We should explore a bailout for the news media industry, although the challenge will be how to reinforce the social center without imposing an unconstitutional ideological test. ?This tactic would have been more effective 10 years ago, when there was still a reasonably robust “mainstream” media to support. We can also provide seed funding for private sector efforts to develop new social networking technologies that are not as structurally suitable for extremist exploitation. My team can brief you as early as Tuesday on some preliminary options in this space and ideas on how to frame this initiative for an anxious American public.

_______________________________________________________________________________

That was written seven years ago, and much has changed. We’ve seen plenty of examples of the destabilizing power of social media, amplified by recent policy changes at Twitter, but we have not yet seen a worst-case scenario such as that described above. Which is not to say we haven’t seen elements play out. This week, for instance, Hamas put out a call for a day of protests, which right-wing websites construed as a “day of jihad,” desperately scouring the newswires for any evidence of an outpouring of violence and finding two examples, in France and China, which may be connected to the situation in Gaza. Neither had been conclusively linked as of this writing, although the China incident looks likely to be.

Even if both of the incidents linked above turn out to be directly or indirectly inspired by Hamas and the unfolding war in Gaza, it’s still a far cry from the scenario in my vignette. More in line with my scenario, we have already seen some hate incidents affecting both Jews and Palestinians outside of the war zone, but not at scale.

Overestimating risk is an occupational hazard, which I have talked about in public a few times. It comes from thinking too much about what could happen and listening too much to what extremists say they want to do. I have mostly stopped trying to forecast short- to medium-term risk in any serious way, because subject-matter expertise does not necessarily correlate to forecasting skill.

On the other hand, events are still playing out. Importantly, most of the really disturbing propaganda is yet to come. We have seen significant posting of violent images, but we have not yet seen wide dissemination of the most horrific images that have been reported to exist. If they truly exist, and I think they probably do, Hamas almost certainly intends to distribute them. And even if they don’t, once the Israelis move into Gaza and casualties start piling up, we’re going to see a lot of gruesome images and video regardless. If you worked OSINT during the Syrian civil war, you know what that means.

But more eyes will be on these images, because of public interest but also because of the new normal on Twitter. While most companies will aggressively work to stem the flow of such material online, Twitter/X can’t be relied on to match the pace, and users are already reporting being bombarded with images of graphic violence.

Which brings us to the second thing that my vignette got wrong. Who needs a White Genocide Reporter app when you have X? While X is still posturing as something other than an overtly white supremacist platform, hardly a week goes by without a new story about the site’s surging hate content. Twitter/X is far bigger than my imaginary worst-case scenario app. In addition to the gutting of Twitter’s trust and safety teams, we’ve also seen Elon Musk excuse the sharing of atrocities when it’s done for the purpose of firing up people on the far right. Even Fortune, a fanzine for billionaires, expects the platform is going to be a “dumpster fire” as the Gaza war continues.

I do think the vignette got one thing right, which was already true when I wrote it, and more so now, we desperately need to rebuild our journalism infrastructure. Today, I am far less optimistic that a stronger fourth estate would fix our hate problem, and I cringe a bit at the “center of society” language I used then. But without reliable, objective information sources, we’re lost, in ways I did and did not anticipate in 2016.

That’s all I have to say about this situation for now. To be honest, I’m mostly overwhelmed with sorrow for the children who will suffer and die on both sides of this conflict, for the families failed by their leaders, for what all of this says about human nature. But I can’t go there right now, not really. So I chip at the edges, in between larger projects that hopefully illuminate some of these problems.

Note: Next week’s newsletter will be written in advance, as I expect to be out of the office, so it won’t be tracking current events. I expect to return to business as usual the following week.