The Guardian's headline tonight: "Netanyahu rejects US-led 15-point Gaza peace plan in rare public break with Trump." * Now look at the address bar of that same story: theguardian.com/world/2026/aug/09/**israel-rejects-us-led-15-point-gaza-peace-plan-says-netanyahu**. * Slugs are generated from the headline at the moment of publication and then frozen, because changing them breaks every shared link. So we can read the first draft off the URL: "Israel rejects US-led 15-point Gaza peace plan, says Netanyahu." Flat. Sourced. Correct. The Guardian's own live blog was still linking to the piece under that wording an hour later. * Somewhere between publishing and now, "says Netanyahu" became "in rare public break with Trump." Four words were added, and they are the only four words in the headline that make an argument. * So let us check the argument. Is it rare, and is it a break? * ### What the Guardian itself reported on July 31 * This is where we have to hand the Guardian its own clipping. Nine days ago, its story on the deal ran under the headline: "Trump says 'complete disarmament' of Hamas agreed in Gaza but Israel yet to accept." * *Israel yet to accept.* In the headline. On day one. * And in the body of that same July 31 piece, an Israeli official told Reuters — quoted in full by the Guardian — that "Israel demands the complete disarmament of Hamas, including the removal of weapons from Gaza and the complete demilitarisation of the strip **as a precondition for any process**," adding that the plan under discussion did not "provide a satisfactory response to these demands and Israel conveyed its objections." * Complete disarmament. As a precondition. Objections conveyed. Published by the Guardian on July 31, 2026. * Tonight Netanyahu told his cabinet the army "will not carry out any withdrawal until Hamas is genuinely disarmed," and added, "We are talking about genuine disarmament, not fictitious disarmament." * Those are the same position, nine days apart. One of them was filed as a caveat in paragraph fourteen. The other is a rare public break. * ### The part almost everyone has already forgotten * Here is the bit that turns this from a headline quibble into something worse, and we can be blunt about it because we wrote it down at the time. * On July 31, the plan said disarmament would be "gradual, sequential and time-bound" and "linked to an Israeli withdrawal, in phases." The Board of Peace's own high representative, Nickolay Mladenov, put it in a sentence you could hang on a wall: "Withdrawal must move in lockstep with decommissioning." * Then on Monday, August 3, Mladenov went to see Netanyahu. And on August 4, the Board of Peace issued this, which Al Jazeera headlined and most newsrooms filed under miscellaneous: "Contrary to inaccurate reports, we note that the withdrawal of the [Israeli military] beyond the Yellow Line will take place only once decommissioning is complete... This applies to light weapons, heavy weapons, and the tunnels alike." * Lockstep, meet only-once-complete. The Board of Peace did not clarify its plan on August 4. It swapped the sequencing — the single term the entire deal turns on — into Israel's preferred order, and called the previous week's accurate reporting "inaccurate reports." * So when Netanyahu says tonight that nobody withdraws until Hamas is genuinely disarmed, he is not breaking with the Board of Peace. **He is quoting it.** That is the Board's published position, five days old, issued from a room he was sitting in. * ### Then what exactly did he reject? * Something real, to be fair, and we should say so plainly rather than pretend the story is nothing. * The 15-point roadmap published on July 31 does still exist on paper, and on paper it is phased. NPR's write-up tonight was the only one we read that told you this outright — that the roadmap "envisioned Israel's withdrawal from Gaza and Hamas's disarmament as part of a phased process, rather than requiring Hamas to fully disarm before Israeli forces began pulling back." Against that document, Netanyahu's statement is a rejection. Credit where it is due: NPR's readers finished the story knowing which version of the plan was being refused. Nobody else's did. * Netanyahu also went wider than Gaza, saying he was facing a "wave of rumours about Israeli weakness and Israeli withdrawals on all fronts" and rebuking US pressure to end the wars in Gaza, Lebanon and Iran without Israel's buy-in. That is genuine public friction with Washington, and the Guardian's standfirst carried it accurately. * But that is a story about a plan with two contradictory sequencings, both written by the same board, nine days apart. It is not a story about a man who surprised anyone. * ### Who this framing serves * Score the two versions against each other. * "Israel rejects US-led 15-point Gaza peace plan, says Netanyahu" points at a policy and a government. You would go looking for the plan. * "Netanyahu rejects US-led 15-point Gaza peace plan in rare public break with Trump" points at two men and a relationship. You would go looking for the feud. * The second one is a better read and a worse map. It puts Trump at the centre of a story whose actual centre is a demarcation line called the Yellow Line, behind which Israeli forces still sit on roughly 70% of Gaza, according to Palestinian officials and aid agencies cited by NPR. And it quietly relocates the news from *what changed on August 4* to *how Trump must be feeling on August 9* — which has the convenient property of requiring no archive. * Meanwhile Hamas said Sunday it remains committed "to the roadmap agreed upon," and its political bureau member Bassem Naim asked the mediators to "pressure Netanyahu and his government to adhere to the roadmap and not obstruct the process for internal political and electoral reasons." Both sides now claim to be the ones honouring a document that has said two different things about who moves first. * Nobody moves. Everybody is compliant. The roadmap is unfalsifiable, which is a lovely property for a roadmap to have if what you actually need is time. * ### The tell * We keep coming back to that URL, because it is the cleanest evidence in the whole story. * Somebody at the Guardian wrote the accurate headline first. The reporting underneath it is solid — Andrew Roth's piece has the quotes, the rebuke, the context. Then the headline got a promotion, from what happened to what it means for Donald Trump, and the URL was left behind holding the receipt. * It is a small edit. It took maybe fifteen seconds. And it is the difference between a reader who goes looking for what the Board of Peace published on August 4, and a reader who goes looking for a fight between two politicians. * Only one of those readers finds out that the plan already changed.