The Ceuta trade on Instagram, Russian propaganda laundering in Greek media, and Bolsonaro on screen from house arrest

The Ceuta crossing is being sold on Instagram, for a few hundred pounds

Ceuta is a Spanish enclave on the North African coast, separated from Morocco by a border fence that ends in the sea – and last month roughly 78,000 migrants arrived there by sea in a matter of hours. A BBC investigation published on August 14 looked at who keeps the crossings going: dozens of interconnected Instagram profiles that portray the short swim around the border fence as safe and easy, and offer help with it for a fee.

A BBC reporter set up a dummy account, “Mohammed”, posing as a young Moroccan hoping to reach Europe. One profile replied claiming to have helped “many people”, offered crossing advice for hundreds of pounds, and asked for payment by bank transfer through a named money service. It did not answer whether the money would be refunded if the crossing failed, and later deleted several of its messages. The same profile had told the BBC's real account that it does not encourage immigration to Ceuta – a line several other profiles repeated.

Posts advertising another attempt at the crossing the following weekend were still being actively recommended by Instagram's algorithm when the BBC published. Spanish intelligence, cited by La Razón, has described the mobilisation behind July's surge as decentralised: separate accounts and groups in Morocco, Algeria, Tunisia, France and the US split the work of producing, amplifying and coordinating content, Al Jazeera reported. Several of the profiles the BBC identified are no longer on Instagram.

Russian propaganda is being laundered into Greece's mainstream media

Last month a story swept Greek online media claiming Volodymyr Zelensky's parents were under investigation by NABU, Ukraine's anti-corruption bureau, after $11 million was found in their bank accounts. The story was false: no credible outlet and no Ukrainian authority had reported any such investigation, Greek fact-checking organisation Ellinika Hoaxes found on July 10. An investigation published by Balkan Insight on August 19 documents the route the story took into Greek news: “disinformation laundering”, where fabrications seeded by pro-Kremlin networks travel from fringe sites into popular mainstream outlets.

The claim rested on a video posted on X by the account @SprinterPress, dressed up as “Europe Today”, a programme from Euronews, with a voice added using AI; Euronews had broadcast no such thing, and the logo's colours were inverted from the real one. From there the story was republished across five Greek websites – among them Pronews, a fringe pro-Russian site, and Newsbomb, one of Greece's most popular Greek-language news sites. Newsbomb belongs to DPG Digital Media Group, owned by businessman and Panathinaikos football club owner Dimitris Giannakopoulos, whom BIRN describes as sympathetic to Vladimir Putin.

Ellinika Hoaxes and Ukraine's Centre for Countering Disinformation traced the fabrication to Matryoshka, also known as Operation Overload – a coordinated Russian propaganda network that spreads fake news, deepfakes and memes dressed in the logos of well-known Western media. Russian state outlets such as RT and Sputnik have been banned in Greece under EU sanctions since 2022; the fabricated stories reached Greek readers through Greek-owned sites instead.

An AI Bolsonaro opened Brazil's campaign – now the court may ban deepfakes entirely

In July, Brazil's PL party held the convention that confirmed Flávio Bolsonaro as its presidential candidate. On stage, the campaign played an AI-generated video of his father, former president Jair Bolsonaro. Jair cannot take part in the campaign himself: he is barred from running, sits under house arrest with a 27-year sentence in the coup-plot case, and is banned from publishing political messages, even through others.

The PT, the party of President Lula, who launched his re-election bid on August 16, took the video to the Superior Electoral Court (TSE), which runs Brazil's elections. It calls the video early campaigning and a way around Jair's restrictions. The two sides read the rules differently. Flávio's lawyers say using AI needs no permission from the person shown, and that a deepfake is only banned when it is made to deceive. The PT points to the court's own 2024 rules, which prohibit synthetic content that replaces a living person's image or voice – with or without permission.

On Tuesday, August 18, TSE president Kassio Nunes Marques called the ministers into a closed-door meeting to agree one rule for the whole campaign: whether deepfakes should be banned outright, and how to punish violations – warnings and fines are on the table, more for repeat offences. The meeting ended without a decision, O Tempo reported. The court's direction is visible in three parallel cases, where a majority of ministers has already voted that a deepfake can be taken down without proof it could deceive voters, O Globo reported.

YouTube, meanwhile, showed the TSE a likeness-detection tool: officials, journalists and candidates in a pilot group can register their face and ask for the removal of AI videos that use it, Valor reported on August 11.

Russia's anti-Ukrainian disinformation in Poland surged more than 250%

After a public dispute opened between Warsaw and Kyiv this summer, the amount of anti-Ukrainian content in Polish on social media rose by more than 250%, Polish officials told Reuters on August 14. Deputy digital affairs minister Paweł Olszewski said Russian disinformation “exploded” on social media and official Russian platforms once the dispute began, and that some of the narratives were later picked up by Polish media outlets. The Polish Institute of International Affairs recorded the increase across X, Facebook, Instagram, TikTok and YouTube.

The trigger was a dispute between the two governments. In late May, President Volodymyr Zelensky gave one of Ukraine's special forces units the honorary name “Heroes of the UPA” – the WWII-era Ukrainian Insurgent Army, which Poland holds responsible for the 1943–45 Volyn massacres, in which Poland says around 100,000 Poles were killed, and which it officially recognises as genocide. In June, Polish President Karol Nawrocki stripped Zelensky of the Order of the White Eagle, Poland's highest state honour. The campaign works with that quarrel and with public frustration over Ukrainian refugees, as polls show far-right support growing before Poland's 2027 parliamentary elections.

The method may be new, analysts told Reuters. Much of the increase appeared to be artificially generated by bot activity designed to create the impression of rapidly growing anti-Ukrainian sentiment, PISM analyst Filip Bryjka said, and the messaging focused on amplifying a real controversy rather than spreading invented stories – which makes it harder to detect. Ukraine's Centre for Countering Disinformation, reporting on the same campaign on August 17, added that AI-generated anti-Ukrainian content is circulating on Polish social media alongside outright fabrications – among them a “Ukrainian missile” story and a claim that Poland would host Ukrainian troops for ten years. The flow of disinformation around the dispute, Olszewski said, found a ready audience.