Monthly News Recap, March 2026: The Month They Ran Out of Subtlety
March produced forty-four stories, stacking up fast enough that by mid-month it stopped feeling like news and started feeling like a policy production line. Iran killed the internet during US and Israeli strikes. Russia cut mobile internet, blocked Telegram early, and then arrested twenty people for protesting the block. Meta buried Instagram's encryption removal in a four-year-old support article, while Apple snuck age verification onto UK iPhones without a word of announcement. As you probably guessed, this was not a slow month.
This monthly recap covers all forty-four articles published between March 1 and March 31, 2026. The tools change, the excuses rotate between "child safety," "national security," and "platform accountability," and the direction of travel stays perfectly consistent.
Pentagon Pressures Anthropic to Drop AI Safeguards: What's at Stake?
The US Department of Defense pressured Anthropic to weaken its AI safety commitments as part of a defense contract negotiation, reportedly seeking fewer restrictions on military applications of Claude. Trading away safety standards in a procurement deal means the people writing those standards aren't the ones taking the risks, and someone always ends up bearing the consequences of that gap.
Chinese Comedian's Innocent Joke About Family Reality Ends With Her Social Media Ban
A Chinese comedian made a mild observation about family life and found herself banned from every major domestic platform in the country within days. No political content, no named officials, just a comment that gestured at a reality the government would prefer left unacknowledged. Inconvenience is all that China's censorship infrastructure needs to pull the trigger.
Pay Attention: The UK's Under-16 Social Media Ban Won't Just Affect Kids
The UK launched a public consultation on banning under-16s from social media, but it’s not only children who are affected. Any age-gating system that excludes minors requires identifying everyone else. The infrastructure being built to protect children is also infrastructure built to verify and track adults, and the question isn't about teenagers but about what the verification layer gets used for once it's normalized.
Iran's Regime Pulls the Internet Plug Yet Again Amidst US-Israeli Attacks
As US and Israeli strikes landed in Iran, the regime's first domestic response was cutting internet access while people were trying to reach family and find safety information. It's a pattern Iran runs repeatedly. The moment information becomes most critical, the state removes it. Yet, what's notable is the speed, as the shutdown wasn't measured policy but a reflexive move, which tells you something about how the regime understands the internet.
Australia's War on Privacy Continues, and Now It's Coming for AI, Search, and App Stores
Australia's eSafety regulator threatened to force app stores and search engines to block AI services that missed its March 9 age-verification deadline. A Reuters review found more than half of the 50 most popular AI tools had done nothing visible to comply. Apple pledged "reasonable methods" without specifying what those are, while Google declined to comment. Critics noted none of this stops anyone who knows how to type a URL.
Microsoft Gets Called "Microslop" Due to AI Push, Bans the Term and Locks Its Own Discord
Microsoft banned the term "Microslop" from its official Discord server after users applied it to the company's aggressive AI rollouts, then locked the channel entirely. The sensitivity is perhaps more revealing than the nickname. Banning a word from your own community forum doesn't make the word go away. It makes the forum go away, which tends to confirm whatever grievance generated the insult.
Hundreds of Scientists Just Pleaded Governments to Pump the Brakes on Age Verification
Over four hundred researchers signed an open letter asking governments to pause age verification mandates, flagging that verification requires data collection, collected data gets breached, and the children these laws target are generally the fastest to circumvent them. The governments receiving this letter proceeded to pass more age verification laws anyway, because the point was never really the evidence.
Want to Look Professional? Say Bye to Your Privacy: LinkedIn Identity Verification
LinkedIn rolled out identity verification framed as a trust tool, which glosses over what it involves. Submitting government-issued ID to a third-party provider creates a dataset combining professional networking history with verified real-world identity, considerably more useful to data brokers and state actors than either alone. The cost of looking credible on LinkedIn is handing your real identity to a platform with a documented history of surveillance partnerships.
UK Wants to Ban VPNs to Stop Kids From Avoiding Social Media Bans: What You Need to Know
UK officials floated restricting VPN access to enforce the proposed social media ban for under-16s. The logic is that children will use VPNs to circumvent age verification, therefore, VPNs are the problem. The actual logic runs the other way: If your enforcement mechanism is defeated by a free app that takes thirty seconds to install, the enforcement mechanism is the problem.
TikTok Flat-Out Refuses DM Encryption, Plays the Same Old "Child Safety" Tune
TikTok confirmed it has no plans to introduce end-to-end encryption for direct messages, citing child safety. Unencrypted messages are accessible to TikTok, ByteDance, and any government making a legal request, which for a Chinese-owned platform includes the Chinese government under its national security laws. Child safety is the reason given, but in reality, surveillance compatibility is the result delivered.
Open Source Developers Are Considering Locking California Out, and the State Has Only Itself to Blame
California's AB 1043 requires every operating system to collect user ages and expose that data to app developers via a real-time API, covering Linux distros, SteamOS, and niche projects qualifying as an OS under the law's definition. MidnightBSD updated its license to exclude California users. DB48X, an open-source calculator that apparently qualifies as an OS, banned California and Colorado users, too. When your age verification law has angered the math people, something has gone wrong.
Prosecutors Confirm an Italian Journalist Was Hacked with Paragon Spyware
Italian prosecutors confirmed a journalist was targeted with Paragon's Graphite spyware, adding a verified case to the growing list of EU journalist surveillance incidents. Paragon sells only to vetted government clients, so the question is not whether this was state-sponsored but which state. The EU's existing legal frameworks have so far proven insufficient to stop member states from deploying commercial spyware on journalists within their own borders.
Apple Asks UK iOS 26.4 Beta Users to Verify Their Ages, Admits Screwup
UK users on the iOS 26.4 beta received unexpected age verification prompts, which Apple attributed to a misconfiguration and said had been disabled. "We turned it on by accident" is an uncomfortable explanation for a privacy-invasive feature appearing without warning. Moreover, the infrastructure required to accidentally deploy age verification is infrastructure that has been deliberately built, and accidents tend to reveal exactly what's already ready to run.
Indonesia Bans Social Media for Under 16s, Calls It a "Digital Emergency"
Indonesia passed a social media ban for under-16s framed as a "digital emergency." The enforcement gap is significant. Indonesia has no functional age verification infrastructure at scale, the platforms are global, and affected children are already using VPNs in large numbers. The framing positions any objection as endangering children rather than questioning whether the proposed solution addresses the actual problem, which it doesn't.
Australia's New Online Age Checks: Awkward for Users, Risky for Privacy
Australia mandated age verification for adult content websites, with "meaningful verification" meaning government-issued ID, facial recognition, or credit card details. The data being collected is sensitive, the activity being verified is exactly what people most want to keep private, and the verifying systems are third-party contractors with their own retention policies. Teenagers already know how to circumvent these, likely pushing them toward less regulated alternatives with no safeguards at all.
OpenAI's Questionable Pentagon Deal Just Cost the Company Its Robotics Chief
OpenAI's head of robotics resigned following the company's Pentagon contract announcement, adding to a pattern of senior technical staff departing over the company's growing willingness to take defense work. The contract covers AI for military logistics and operations planning, and when the people actually building the technology start leaving over where it's being pointed, that says something about OpenAI's stated safety commitments that no press release will.
Platforms That Can't Stop Harassment Have Somehow Mastered Censoring Women's Health
Social media platforms consistently suppress content related to women's health, including menstruation, pregnancy, and reproductive medicine, while leaving targeted harassment campaigns significantly less aggressively moderated. Platforms that claim content moderation is technically difficult have built systems that are selectively very good at it — what gets moderated versus what doesn't is a revealed priority, and it tells you more about a platform's actual values than any community guidelines document ever will.
UK MPs Reject a Blanket Social Media Ban for Under-16s in Favor of Consultation
The UK House of Commons voted against a blanket social media ban for under-16s, opting for public consultation instead. But while this is the right outcome, it’s mostly for the wrong reasons. The MPs who voted against it mostly preferred a different enforcement mechanism, not concerns about surveillance implications, and the consultation will almost certainly produce a version of the same policy, structural problems and all.
The Silent Backbone of the Internet, and Why Governments Are Eyeing It
An examination of undersea internet cables, the physical infrastructure carrying roughly 95% of international internet traffic, and why military and government interest in them has intensified alongside rising geopolitical tensions. From documented damage incidents to surveillance taps and legal protection frameworks that are substantially less robust than the infrastructure they're meant to cover – the internet feels weightless, but the cables carrying it are physical, cuttable, and increasingly contested.
Studying the Internet in the US Can Now Get You Deported
A researcher studying internet freedom was deported from the United States, joining a pattern of academics whose work touching on surveillance or digital rights has resulted in deportations under the current administration. Research requiring US-based access to infrastructure or datasets is now being conducted under the assumption that the researcher's continued presence in the country depends on their findings not being inconvenient.
Meghalaya Cut Off Six Districts' Internet After Election Violence Turned Deadly
Indian state authorities shut down internet access across six Meghalaya districts after election violence turned fatal, framing it as a measure against misinformation. Cutting internet access also prevents journalists from reporting, citizens from documenting, and people from reaching emergency services, all of which weigh heavily against the claimed benefit, particularly when the government controls most of the information environment during the shutdown anyway.
Manila Councilor Accused of Lascivious Conduct Had a Reporter Arrested for Covering It
A Manila city councilor facing lascivious conduct accusations had a reporter arrested for publishing a court-issued warrant, going with cybercrime law as the pretext. Using cybercrime legislation to punish the publication of a public legal document is a significant stretch of the statute and one that conveniently silences coverage of a public official's misconduct. The reporter was covering a warrant, and the councilor is the one with something to hide.
Eleven African Governments Built a $2 Billion Chinese Surveillance Grid
Eleven African governments collectively built $2 billion worth of Chinese-supplied surveillance infrastructure covering CCTV networks, facial recognition, and communications interception, financed largely through Chinese development loans that run alongside the surveillance relationship. Framing this as a developing-world concern misreads both the technology and the political logic driving its adoption. The same systems, the same financing model, and the same debt-surveillance linkage are already spreading well beyond Africa.
Congo Cuts Internet During Presidential Elections and Calls It Democracy
The Republic of Congo shut down internet access during its presidential election, describing it as protecting the integrity of the vote. Cutting the internet prevents independent monitoring, prevents citizens from documenting irregularities, prevents opposition candidates from communicating with supporters, and prevents international observers from reporting in real time. These are very clearly not side effects, and calling them democracy is the audacious part.
Moscow Is Back to Pagers and Paper Maps After Russia Killed Its Mobile Internet
Russia ran a mobile internet blackout across Moscow described as a security drill, leaving residents unable to use maps, messaging, or mobile payments for hours. But while the official version is drone defense, things start to look differently when you learn that Russia has invested heavily in TSPU deep-packet-inspection hardware specifically to make shutdowns possible, and running them periodically tests the infrastructure while normalizing internet absence before it's needed in earnest.
Cloudflare Stands Up for the Open Internet Against Italy's "Piracy Shield"
Italy's Piracy Shield was issuing blocking orders so broad they were taking down legitimate services alongside infringing ones, including Cloudflare infrastructure used by hundreds of unrelated websites. Cloudflare pushed back publicly, publishing the orders and calling out the collateral damage. One of the few cases this month where a major platform did the right thing under pressure, though Cloudflare's motivation is partly self-interested, given that its own infrastructure is what's being blocked.
Meta Removes Instagram Encrypted Chats by Updating a Four-Year-Old Article
Meta removed end-to-end encryption from Instagram direct messages and disclosed it by quietly updating a support article from 2021. There was no announcement or even a notification to users. Burying a significant privacy downgrade inside a routine help-center edit is a specific choice about how much users deserve to know about changes to their own communications. The answer Meta demonstrated is not much, and ideally nothing at all.
Defeating InfoStealer Malware
An in-depth guide from contributor Jesse William McGraw on infostealer malware, covering what it is, how it operates, what it steals, and how to defend against it. Infostealers are among the most financially motivated malware categories in active deployment, responsible for credential theft at scale and a significant share of initial access cases leading to larger breaches. Written from the perspective of someone who has spent time on the threat actor side.
Brazil's Social Media Ban Is Live, and Brazilians Are Already Circumventing It
Brazil's social media restrictions came into effect, and VPN downloads spiked almost immediately. Governments pass platform restrictions, citizens reach for circumvention tools, and governments then consider restricting those tools. Brazil's ban targets specific platforms rather than content categories, making it narrower than some but no less revealing about what happens when populations decide not to comply with digital restrictions they consider unjust.
Telegram Goes Dark in Russia Two Weeks Early, "Coinciding" With Moscow Blackout
Russia's Telegram block activated roughly two weeks ahead of its announced deadline, timed alongside the Moscow mobile internet blackout, with Roskomnadzor offering no explanation. Moving the block forward without notice demonstrates the state's willingness to act faster than users can prepare, and the coincidence with the Moscow blackout drill suggests both moves were coordinated rather than independent.
China Just Censored a CNN Segment on Chinese Censorship, Live, in Real Time
During a CNN broadcast covering Chinese censorship, the signal was cut in China in real time as the segment aired. A government censoring a report about its own censorship, live, on the record, with no apparent concern about optics, is a government that has stopped worrying about appearances and started optimizing purely for outcomes. The segment was about censorship, and ironically, the censorship of the segment is now part of the story.
Millions Trusted Crime Stoppers With Their Safety, and the Platform Was Quietly Logging Them
A Crime Stoppers hack revealed the platform had been logging user data, including IP addresses and browser fingerprints, despite promising anonymity to tipsters. For people who submitted tips about gang activity, domestic violence, or organized crime expecting identity protection, the breach is a physical safety issue. Building an anonymity tool on top of standard web tracking infrastructure is a fundamental design failure, not a configuration mistake.
Ofcom Fined 4chan £520,000 and Got an AI-Generated Hamster in Return
Ofcom fined 4chan £520,000 for failing to comply with the UK's Online Safety Act transparency requirements, and 4chan responded with an AI-generated image of a giant hamster. The Online Safety Act's enforcement against platforms with no UK commercial presence raises real questions about whether Ofcom can actually compel anything here, and 4chan appears to have reached the same conclusion, and the hamster response made it very clear.
No Archive, No Accountability: The Internet's Memory Is Under Threat
The Internet Archive, the nonprofit behind the Wayback Machine, has been facing legal challenges, funding pressure, and operational strain threatening its long-term viability. Journalists, researchers, lawyers, and ordinary people use it to prove something was said, that a page existed, that a policy was different last year. Losing it doesn't just affect historians. It affects everyone who needs to hold institutions to their previous statements.
Turkey Jails a Journalist Mid-Holiday for Covering Institutional Corruption
Turkish authorities arrested a journalist during the Eid al-Fitr holiday for reporting on institutional corruption, and the timing was absolutely deliberate. Holiday arrests reduce public response and media attention while demonstrating the state's appetite for silencing coverage doesn't pause for public holidays. Turkey's broader pattern of using criminal law to prosecute journalists whose work embarrasses state institutions or officials gets even clearer than before.
GrapheneOS Chooses Privacy Over Compliance, and Accepts the Consequences
GrapheneOS publicly committed to maintaining its hardened privacy architecture even where that means declining to comply with regulatory frameworks requiring it to be weakened. A privacy operating system that installs backdoors on regulatory demand is no longer a privacy operating system. GrapheneOS's value is its refusal to treat compliance as a higher priority than the security guarantees it makes to its users, with the real costs acknowledged plainly.
A New Mexico Jury Just Found Meta Liable for Lying About Child Safety
A New Mexico jury found Meta liable for misrepresenting its child safety practices, finding the company made false claims about protections in place for minors. The verdict establishes that courts are willing to treat corporate child safety claims as factual assertions subject to liability, not marketing language immune from scrutiny. Meta said children were protected when they demonstrably weren't, and New Mexico's jury treated that as a misrepresentation.
The Pentagon Is Running a Coordinated Censorship Operation Around the Iran War
The Pentagon has been running a systematic effort to restrict, delay, and shape media coverage of its military operations in the Iran conflict, using embedding restrictions, satellite communication controls over press corps equipment, and coordinated pressure on news organizations. These are documented mechanisms, specific editorial decisions with traceable Pentagon fingerprints, and restricting coverage of a war you're fighting is censorship regardless of which administrative channel it travels through.
Apple Made Age Checks Mandatory for UK iPhones and Skipped Any Announcement
Apple quietly made age verification mandatory for UK iPhone users with no press release, blog post, or notification to affected users. The silence is deliberate, as Apple has been navigating UK age verification requirements for months, and rolling out compliance changes without announcement is a way of testing the infrastructure while minimizing the public conversation about what it actually collects.
When AI Knows All Your Dirty Secrets
Contributor Jesse William McGraw examines how people actually use AI as a confidant for sensitive information, against what AI companies can see, store, and share. The catalyst is the Tumbler Ridge high school shooting in Canada, where the perpetrator apparently communicated intent to ChatGPT, which flagged but did not escalate externally. The case forces a confrontation with what AI companies know, what they're obligated to do with it, and what privacy is realistic on a monitored platform.
Russia Detained 20 Protesters for Demanding Telegram and the Internet Back
On March 29, rallies against Russia's Telegram ban took place across multiple cities. Authorities rejected all 25 permit applications from organizing group Scarlet Swan, then detained twenty people when rallies happened anyway, including human rights activist Aleksandr Podrabinek and at least two minors, with mobile signals being jammed in Murmansk ahead of the rally. The Kremlin wants Russians on MAX despite officials being too wary of FSB surveillance access to use it themselves.
Austria's Social Media Ban Follows a Tired but Slightly Improved Playbook
Austria announced a social media ban for under-14s following Australia's enforcement model, with draft legislation due by June. The ban will work about as well as every other ban but what sets the announcement apart is a compulsory new school subject called "Media and Democracy," teaching students to identify disinformation and recognize manipulation. That's the only part worth exporting to the rest of Europe, and none of them have announced anything like it.
What's Really Running Inside Your Free VPN: A Mysterium VPN Research
Mysterium VPN ran static analysis on 18 popular free Android VPN apps, examining trackers, permissions, hardcoded endpoints, and connections to OFAC-sanctioned jurisdictions. Seventeen of 18 contained at least one embedded tracker, averaging 4.9 each. Six communicated with OFAC-flagged domains and five connected to Chinese servers. Some communicated with 105 unique hardcoded domains, while others requested numerous dangerous permissions, proving that free VPNs are not the privacy and security tools that they pretend to be.
Australia's Social Media Ban Corners Big Tech While Privacy Stays at Risk
Over three months after Australia's under-16 social media ban took effect, eSafety opened formal investigations into five major platforms for non-compliance. Surveys found roughly seven in ten previously registered children still had accounts, with platforms letting underage users retry age checks until they passed. The five face fines of up to $49.5 million. The harder part is that verifying who is under 16 requires verifying everyone, meaning identity checks for every user on a mandate that keeps expanding.
Be part of the resistance, quietly.
Get Mysterium VPN
Dominykas is a technical writer with a mission to bring you information that will help you in keeping your digital privacy and security protected at all times. If there's knowledge that can help keep you safe online, Dominykas will be there to cover it.
Related Articles
.png)
.png){kind=small}
