About this issue. It covered the week ending 22 September 2026 and was held back before launch. It was published on 25 September 2026 with its counts rebuilt under the current method, so the scorecard at its foot reads as the method now counts, not as it counted when the issue was written. The threads it opened are followed in the issues after it.
What moved
Russia's governing party takes a record majority, and independent analysts dispute the count
United Russia will hold 355 of the 450 seats in the State Duma, Russia's lower house, up from 315, according to preliminary results reported by Folha de S.Paulo and g1. The previous record was 343 seats in 2016. Russia's population was 143.7 million in 2024 (World Bank), so this is the legislature for roughly one person in every fifty alive.
The vote ran 18–20 September and was the first parliamentary election since Russia's full-scale invasion of Ukraine in 2022. Russia's Central Election Commission put United Russia's party-list share at 57.96% with 97.37% of precincts counted, and turnout above 59% (Meduza).
Two independent estimates contradict those figures, and this is the part worth carrying. Mathematician Boris Ovchinnikov and election observation expert Roman Udot, speaking to the broadcaster TV Rain, put actual turnout at about 40% rather than 59%. Separately, Novaya Gazeta Europe calculated that without ballot stuffing United Russia would have taken about 34% of the vote rather than 58% — roughly 1.7 times its real support — using a method developed by the analyst Sergey Shpilkin, with two other analysts agreeing with the calculation (Meduza). Shpilkin's method compares vote share against turnout across polling stations; the outlet notes the method has limitations. We have not seen the underlying data.
Analysts also describe a specific intervention. The party A Just Russia was sitting near the 5% threshold needed for seats when roughly 60,000 votes appeared for it around midday on 21 September; one expert counted 67,330 votes added in a single hour. It finished at about 5.1% and will be in the Duma. "It's just one phone call, the decision is made, they rewrote the number right there," the electoral analyst Stanislav Andreychuk told Carnegie Politika (Meduza).
Forty-nine of the new deputies fought in Ukraine, about 11% of the chamber, according to United Russia's secretary-general Vladimir Yakushev — who first said 86 and later corrected the figure (g1). Yabloko, the only registered party opposing the war, was barred from the ballot by the courts. Kremlin spokesman Dmitry Peskov said the result showed "the highest level of cohesion of Russian society around the head of state"; the European Commission called it "another major setback for Russian democracy" and described "systematic and institutionalised repression". Voting was also held in occupied parts of Ukraine, where the annexations are not internationally recognised.
Our sources disagree on the smaller parties. g1 reports the Communist Party on 40 seats, the LDPR on 25 and New People on 20. Folha reports the Communists falling from 57 to 40 and New People falling from 15 to 4. Both cite preliminary Central Election Commission data; the final result was due on 25 September. We cannot resolve this, so we are not picking one.
One footnote that sits oddly beside the turnout claim: among Russians who voted outside the country, United Russia came first at only 9 of the 27 such polling stations analysed by the investigative outlet Agentstvo, taking 39.58% against New People's 24.49% (Meduza).
Iran offers to reopen the Strait of Hormuz within seven days
A shipping lane that much of the world's seaborne energy passes through has been closed for most of this year, and this week there was movement toward reopening it.
A senior Iranian official told Reuters, and separately Kyodo News, that the strait could be open again a week after Washington moves. The conditions named were a reduction in military pressure and an end to the blockade of Iran's ports. The proposal reached Washington through mediators on 16 September (The Straits Times, citing Reuters). "The US needs to announce that it wants to resolve the issue diplomatically, make that official, and then agree on a timeline," the official said. Iran's President Masoud Pezeshkian travelled to the UN General Assembly but will not meet President Trump there.
The closure follows a war the United States and Israel launched against Iran on 28 February, which Tehran says killed thousands of people. Iran's answer was missiles and drones fired at targets around the region, and the closing of the strait. A memorandum reached in June was meant to produce a peace deal within 60 days and collapsed within weeks (Anadolu Agency).
Two other conversations pointed the same way this week. Italy's foreign minister Antonio Tajani said he discussed reopening Hormuz with his Iranian counterpart Abbas Araghchi at the General Assembly, and that Italy would keep supporting efforts to restore freedom of navigation. And in a phone call on 22 September, President Putin and Saudi Crown Prince Mohammed bin Salman emphasised "the safe and unimpeded passage of vessels" through both Hormuz and Bab el-Mandeb, and urged a swift end to hostilities in Yemen (The Straits Times, citing Reuters and the Kremlin).
Saudi Arabia has a direct stake: with both Hormuz and Bab el-Mandeb disrupted — the latter by Iran-allied Houthi forces — it is exporting oil through Egypt's Suez Canal instead. The United States bombed the Houthis for two months last year and has so far refused Mohammed bin Salman's repeated requests to rejoin that fight. Reuters reports, citing the New York Times, that Trump called off strikes on the Houthis on Sunday at the last minute, after bombs had already been loaded onto planes. Iran's military central command said on Sunday it had been told the United States was preparing to restart operations with regional support, and warned it would retaliate "without limitations and considerations".
We are not putting a figure on how much of the world's oil moves through Hormuz. Every source we read this week described it qualitatively — "a key maritime route for global energy and commercial shipping" — and none gave a number we could cite.
The United States, Denmark and Greenland sign an Arctic security agreement
Greenland's population was 56,836 in 2024 (World Bank) — smaller than a mid-sized town almost anywhere else. It is now the subject of a signed security treaty with the United States.
The signing took place on Tuesday at the UN General Assembly, with Trump joined by Mette Frederiksen, Denmark's prime minister, and Jens-Frederik Nielsen, Greenland's. Under terms published by the Danish Foreign Ministry, the United States may enlarge and modernise its existing Pituffik Space Base and establish new defence areas at Narsarsuaq and Mestersvig; it gains "free access to and movement between" defence areas "including the Territorial Waters, by land, air and sea", and its aircraft may fly over and land anywhere on the island. The text also restates Danish sovereignty, reaffirming what it calls "the sovereignty and territorial integrity of the Kingdom of Denmark"; Greenland's "right to self determination" is recognised in the same clause (Euronews).
Trump told the General Assembly the agreement gives the United States "permanent control over security and all other needs" of the territory, and that no US adversary may establish a military presence or make sensitive investments there without Washington's "express written approval". He said two "very major military bases" would be built (Anadolu Agency).
The two new sites are worth naming precisely, because they show what the agreement physically involves. Narsarsuaq in southern Greenland is a Cold War base the United States closed in the 1950s, and its population has since fallen to a few dozen people. Mestersvig on the east coast is an outpost that Denmark's Sirius Dog Sled Patrol, a special forces unit, uses today (Kathmandu Post, citing Reuters).
Denmark's own framing differs from Trump's. Copenhagen says the agreement places Arctic security under NATO's watch rather than leaving it to the United States and Denmark alone. Frederiksen said it "strengthens our common security in the Arctic and the North Atlantic Area". Both accounts describe the same document; they emphasise different things about it, and the full text was not public before signing.
This settles, for now, a confrontation that ran for months. Trump had said the United States intended to annex the territory and repeatedly declined to rule out using force. Tensions peaked in January, when European members of NATO deployed additional troops to the island; after talks with NATO Secretary General Mark Rutte, Trump said he would not invade and withdrew a threatened 10% tariff on eight European countries that had objected (Al Jazeera). Polling cited by both Al Jazeera and Reuters shows a large majority of Greenlanders reject US control. We did not find polling published this week, and we have not seen the questions or sample.
The slow story
Nothing this issue. The indicator watchlist is empty because this is the first issue and the backlog has not been built; the only data pulled this week were the population figures cited above. Rather than assemble a trend from a single week of headlines, we are leaving the slot empty and saying so.
Still happening
Nothing this issue. There are no past issues to follow up. This section starts working from issue two.
Pattern watch
Four threads open with this issue.
- Arctic militarisation. Opened 22 September 2026. Two new US defence areas agreed in Greenland, with the treaty text published by Denmark after signing.
- Hormuz and Bab el-Mandeb. Opened 22 September 2026. Hormuz closed since roughly March; Iran offering a seven-day reopening against US steps; Bab el-Mandeb contested.
- Russia's political system after the war. Opened 22 September 2026. Record Duma majority, the only anti-war party barred, independent analysts disputing turnout and vote share.
- Yemen displacement. Opened 22 September 2026, not yet written up. The UN warned of movement toward Djibouti; the story scored well but we did not research it this week.
What this issue does not contain, and why
Two stories scored as highly as the ones above and did not run.
A UN experts' finding that Nicaragua's crackdown on opponents amounts to crimes against humanity had the widest gap in this week's pool between how much it matters and how little it was covered — and we held it, because only two newsrooms in our sources carried it and our own clustering had mixed it up with a separate finding about US strikes on alleged drug boats. Two sources is below the floor we set for ourselves. It is first in line next week.
The run-up to a US–China summit also scored highly, and we left it because what we had was three different stories about one meeting rather than one development. The summit itself is the thing to cover.
We also noticed, without being able to measure it properly yet, that the story our sources covered most heavily this week was a dispute over White House press access, carried by eight newsrooms based in six regions. The Nicaragua finding was carried by two. We are not yet able to put numbers on that kind of mismatch; the instrument for it is not built.
How this issue was made
Researched and written by an AI system working to a published rubric, with every source linked. Candidates came from a standing list of 54 news feeds across ten world regions, chosen to weight coverage toward where people actually live rather than toward where the most news is published; 49 of those feeds produced anything this week. Their 1,630 items produced 22 candidates, which the scoring session merged to 19 stories where several clusters were one development; those were scored on scale, durability, trend signal, attention gap and corroboration, and three were researched and written up. Every candidate's marks are at the foot of this issue. Every figure above was fetched and read during this run — nothing is quoted from memory — and where sources disagree, we have said so instead of choosing.
This is the first issue. Known problems with how it was produced are recorded in the run notes and on the how stories are chosen page, including that our clustering splits a story into several when outlets write in different languages: the Russian election above reached us only through Portuguese and Spanish sources, and the independent analysis that makes it a story had to be fetched deliberately because the software did not connect it.
Corrections
None yet.