https://jolek78.writeas.com/tag:netdelusion thoughts from a friendly human being Mon, 11 May 2026 04:29:28 +0000 https://i.snap.as/DEj7yFm4.png https://jolek78.writeas.com/tag:netdelusion https://jolek78.writeas.com/iran-2026-17-years-later-same-mistake?pk_campaign=rss-feed <![CDATA[It was a Saturday in 2015, perhaps 2016. I was still "normal" back then, still convinced that technology was inherently positive, potentially revolutionary, still naive enough to believe that the internet liberated by definition. I was browsing books at Waterstones on Sauchiehall Street in Glasgow—one of my little guilty pleasures since I landed in Scotland—when I came across "The Net Delusion: The Dark Side of Internet Freedom" by Evgeny Morozov. I picked up the book, went downstairs, sat in the in-house café and started reading. And I went into crisis. His thesis demolished, piece by piece, the narrative of the "Twitter Revolution" of 2009 in Iran. In the book, Morozov cited an analysis by Golnaz Esfandiari, an Iranian journalist for Foreign Policy, who had done something simple but, these days, almost revolutionary: journalism (if you're laughing at this point, you're good people...). She had looked at where the tweets with #iranelection actually came from during the 2009 protests. And the answer? From the West. Not from Iran. Wait, what? Yes, exactly. It was theater. Western self-celebration masquerading as solidarity. !--more-- I remember closing the book with an unpleasant feeling. Morozov doesn't give you the satisfaction of choosing a side in history. He forces you to see that technology amplifies everything—the good and the bad, freedom and control. And that authoritarian regimes have a very steep learning curve, unfortunately. Fifteen years later, the young people in Tehran are trying again: they're taking to the streets trying to overthrow the regime. In the West, I thought we had learned our lesson, that we would stop projecting our technological fantasies onto real protest movements. Obviously, I was wrong. Iran 2009, or when Twitter (didn't) overthrow a regime To understand why Iran 2026 is déjà-vu, we need to go back 17 years. June 2009. Mahmoud Ahmadinejad is re-elected president of Iran with 63% of the vote. The opposition—led by Hossein Mousavi—cries fraud. Millions take to the streets. Tehran fills with green. It's the explosion of the "Green Movement." And here begins the narrative that would define a decade. CNN headlines: "Iran's Twitter Revolution." Time Magazine puts Twitter on the cover with the Iranian flag. Andrew Sullivan—a famous blogger at the time—obsessively tweets using #iranelection and is called "the voice of the Iranian people." Western media cite tweets as if they were dispatches from a war zone. The story was beautiful: young Iranians, tech-savvy and hungry for democracy, were using Twitter to organize protests, coordinate demonstrations, evade regime censorship. Facebook to plan, Twitter to coordinate, YouTube to document. It was the digital revolution overthrowing a dictatorship. Technology defeating repression. The good guys defeating the bad guys. The US State Department was so convinced of Twitter's importance that Jared Cohen—an official—sent an official email to Twitter asking them to "delay scheduled maintenance" so as not to interrupt the Iranian protests. Twitter agreed. Then came Golnaz Esfandiari, an Iranian journalist for Radio Free Europe/Radio Liberty. Where did the tweets actually come from? In June 2010, a year after the protests, Esfandiari published an article in Foreign Policy titled "The Twitter Devolution." She wrote: "Western journalists who couldn't reach—or didn't bother reaching?—people on the ground in Iran simply scrolled through the English-language tweets posted with tag #iranelection. Through it all, no one seemed to wonder why people trying to coordinate protests in Iran would be writing in any language other than Farsi." Question: Why would Iranians organizing protests in Iran write in English? Esfandiari had identified the main Twitter hubs commenting on the Tehran protests and discovered something embarrassing: one was in the United States, one in Turkey, one in Switzerland. The latter's profile stated they "specialized in urging people to take to the streets." She interviewed Mehdi Yahyanejad, manager of Balatarin (one of the most popular Farsi-language websites) who said: "Twitter's impact inside Iran is nil [...] Here [in the United States], there is lots of buzz. But once you look, you see most of it are Americans tweeting among themselves." Iranians—the real ones, in the streets—used SMS, phone calls, word of mouth. Traditional methods. Twitter was mainly useful for one thing: letting the world know what was happening. Documentation, not organization. But the numbers were even worse. In his 2011 book, Morozov cited data that made everything even clearer: only 19,235 Twitter accounts registered in Iran (0.027% of the population) on the eve of the 2009 elections. And many Green Movement sympathizers had changed their Twitter location to "Tehran" to confuse authorities, making it nearly impossible to distinguish whether people tweeting from Iran were in Tehran or, say, Los Angeles. An Al-Jazeera analysis cited by Morozov clarified that fact-checking during the protests had confirmed only 60 active Twitter accounts in Tehran. Sixty. And when Iranian authorities tightened their grip on online communications, that number dropped to six. Vahid Online, a prominent Iranian blogger who was in Tehran during the protests, dismantled the Twitter Revolution thesis even more directly: "Twitter never became very popular in Iran. [But] because the world was watching Iran with such [great interest] during those days, it led many to believe falsely that Iranian people were also getting their news through Twitter." Morozov put it with a perfect metaphor: "If a tree falls in the forest and everyone tweets about it, it may not be the tweets that moved it." At this point in the story one could say "okay, the protests happened in Iran and the West encouraged and celebrated them. What's wrong with that?" Nothing, except that the ayatollah regime learned the lesson. This was the part of Morozov's thesis that had really shaken me. In plain terms, while the West was self-celebrating the "Twitter Revolution," the Iranian government was taking notes. They understood that social media could be more useful to them than to activists. They could track who posted, they could identify protest leaders, they could infiltrate groups, they could use data to arrest, torture, kill. The 2009 protests were brutally suppressed. The Green Movement failed. And the regime emerged stronger, more experienced, more prepared to use technology as a weapon of control. But we were the good guys helping, right? Esfandiari and Morozov tried to tell us we were doing everything wrong, that we were projecting our fantasies and underestimating authoritarian regimes. Did we listen? Evidently not. Iran 2026: same film, different cast December 28, 2025. Protests begin in Iran. Economic crisis, the rial—Iran's currency—collapsed to 1.4 million per dollar, 40% inflation, UN sanctions reimposed in September, the entire Iranian "Axis of Resistance" in tatters after the 12-day war with Israel in June. The streets fill. First Tehran, then the whole country. 31 provinces. Millions of people. And of course, social media explodes. Twitter/X fills with videos, slogans, messages of solidarity. Western media cite tweets as primary sources. Reza Pahlavi—the exiled heir of the Shah deposed in 1979—calls for protests from his social accounts. Persian TV channels in exile (Manoto, Iran International) broadcast 24/7. The US State Department operates a Persian-language Twitter account (@USABehFarsi) constantly posting messages of support. Repetita (non) iuvant. It's 2009 again. Same narrative, same enthusiasm, same conviction that this time—this time for real—Twitter and social media will overthrow the regime. Then, on January 8, 2026, on the twelfth day of protests, the Iranian regime does something interesting. It shuts off the internet. Completely. And I, being a good nerd who doesn't sleep, lives at night, and does things better left unsaid—sorry, the statute of limitations hasn't expired yet—asked myself: wait. If the internet is off in Iran, where is all this content coming from? Who is telling this story? And above all: are we making the same mistake as 2009 again? So, browsing here and there, I came across a long article by Shahram Akbarzadeh—professor of "Middle East & Central Asian Politics" at Deakin University—titled "The web of Big Lies: state-sponsored disinformation in Iran." And I started reading. Before moving forward: stop Let's make one thing clear right away, because I already know someone will misunderstand: I stand in solidarity with those protesting in Iran. Completely. A theocratic regime that kills protesters—estimates range from 44 to 20,000 dead, impossible to know for certain precisely because of the blackout—deserves nothing but condemnation. The reasons for the protests are real, legitimate, understandable. Devastating economic crisis, systematic repression, 47 years of religious dictatorship. Those who take to the streets risk their lives. And they do. But solidarity doesn't mean suspending critical thinking. It doesn't mean uncritically accepting every narrative being sold to us. It doesn't mean ignoring who is constructing this narrative, how and why. On the contrary. If we truly care about the Iranians who are protesting, we have a duty to understand what's really happening. Because wrong narratives have real consequences. And the consequences are always paid by them, not by us tweeting from the couch. So: solidarity yes, but also questions, if no one minds. Technical Box: the evolution of digital censorship TL;DR: Iran didn't simply "pull the plug." It implemented the most sophisticated layered censorship system ever seen, which leaves infrastructure apparently normal while completely isolating the population. It's precision censorship, not sledgehammer censorship. 8:30 PM IRST (5:00 PM UTC). NetBlocks, the organization that monitors global connectivity, registers a sudden collapse: Iran goes from 100% to ~3% connectivity in a few hours. Not just mobile, also landlines, also phones. Calling into Iran from abroad? Impossible. Journalists trying from Dubai can't connect. Families abroad can't reach relatives in Tehran. Total blackout. But there's something curious. BGP routes—the paths that make the internet work—remain visible. Iranian servers continue responding to pings. From outside, the infrastructure looks normal. Cloudflare, IODA (Georgia Tech), all traditional monitoring systems see Iran still "online." Yet user traffic has dropped 97%. How is this possible? To understand what happened on January 8, we need a step back. Iran has developed three generations of shutdowns, each more sophisticated than the last: 2019—Brute Force: During the November 2019 protests (which caused ~1,500 deaths), the regime simply removed BGP routes. It's like ripping out cables: crude, visible, it took 24+ hours to implement because every ISP had to do it manually. Economically devastating—banks stopped, the economy collapsed for six days. 2022—"Digital Curfew": During the Mahsa Amini protests, selective targeting. They shut down cell towers in specific areas, slowed internet during protest hours (4:00-10:00 PM), blocked specific apps (WhatsApp, Instagram). More refined, less expensive. 2025-2026—"Stealth Blackout": The final form. And here it becomes technically fascinating. The current system operates at a single national chokepoint—all Iranian ISPs converge at a few state-controlled exit points. There, a layered system filters everything: Layer 1—DNS Poisoning: Any DNS query for foreign domains gets redirected to 10.10.34.34—a private IP serving a generic block page. You search for google.com? You get an Iranian server saying "domain not found." Layer 2—Protocol Whitelisting: Only three protocols pass: DNS (port 53), HTTP (port 80), HTTPS (port 443). Everything else gets silently dropped. SSH? No. OpenVPN? No. WireGuard? No. Any traditional VPN? No. Zero response, zero error, simply... nothing. Layer 3—Deep Packet Inspection (DPI): The showpiece. System purchased in 2008 from Nokia Siemens Networks, continuously updated. The DPI inspects ALL HTTPS traffic: Reads the SNI (Server Name Indication) field in the TLS handshake Inspects the commonName field in certificates Analyzes HTTP headers (case-sensitive!) Injects TCP RST or HTTP 403 block pages on the fly Selective throttling of encrypted traffic. Practical example: you try to visit Twitter via HTTPS. Your browser starts the TLS handshake. The DPI reads "twitter.com" in the SNI field—which travels in cleartext—and injects a TCP RST. Connection terminated. Twitter's server doesn't even know you tried to connect. Layer 4—National Information Network (NIN): The national Iranian intranet. Domestic services (banking, some state news sites) work perfectly. It's the internet... but only Iranian. The result: From the perspective of BGP routers: everything normal From the perspective of servers: ping responds, infrastructure up From the perspective of users: the internet no longer exists It's genius, in the technical sense of the term. During the June 2025 blackout (during the war with Israel), some tools worked: Psiphon: 1.5 million users maintained (one third of normal base) thanks to multi-protocol design Ceno Browser: decentralized peer-to-peer, from 600 to 8,000 active peers Tor bridges: shot up Starlink: worked... for those who could afford it (hotels, offices, a few privileged people) But in the current January 2026 blackout? Even Starlink has started suffering interference. The regime has learned. And the cost? The impact? Hospitals: booking systems offline Banks: digital transactions blocked Pharmacies: impossible to verify electronic prescriptions Shops: many didn't open (POS not working) The real purpose isn't to stop the economy. It's to stop documentation. It's to obscure the massacres. And Signal? There's an interesting detail completely missing from the 2026 protests narrative, and the silence says a lot. Signal—the encrypted messaging app considered the gold standard for activists and dissidents—is barely mentioned. No articles, no appeals, no campaigns to bypass censorship. Yet Signal had been the weapon of choice during the 2017-2018 protests. "Signal has always been advertised as the go-to application for dissidents or activists to stay secure from any state authority," said Mahsa Alimardani, researcher for Article19, in 2021. But what happened? January 2021, after a massive migration from WhatsApp to Signal, the Iranian government labeled it as "criminal content" and blocked it completely. September 2022, during the Mahsa Amini protests, Signal was still blo