A BBC Eye investigation found Instagram, whose ad review relies primarily on automated technology, running about 30 paid ads promoting child sexual abuse material in India, linking to Telegram channels selling it from 99 rupees. Meta said it removed ads and suspended accounts, though Instagram had told the BBC one reported ad did not breach its standards; India's government summoned company representatives.
A conversation held by Greece's Secretary General of National Security, Thanos Dokos, was the product of a hybrid attack that breached security protocols using highly advanced artificial intelligence, according to informal briefings by government sources. The sources said no confidential or classified information leaked, that a communications security upgrade is under way, and that similar attacks have targeted senior European officials.
Marshall Sheffer, 44, a Highland Middle School teacher in Libertyville, Illinois, faces eight felony child pornography counts after prosecutors alleged he used an AI program to place students' heads on other people's bodies. The case began when students reported he was videotaping them, and prosecutors say his phone held numerous AI-altered images; he surrendered Friday.
Turkish Justice Minister Akın Gürlek said an international fraud network was dismantled after Czech judicial authorities shared evidence: suspects at a call centre in Istanbul's Maslak allegedly used AI-supported software to pose as police, prosecutors, soldiers and bank staff, defrauding Europeans, including Czechs. The centre held 80 foreign suspects; proceeds were allegedly routed through bank accounts and crypto platforms for laundering. Three addresses were raided.
Dutch broadcasters Pointer KRO-NCRV and NOS allege Philip Morris International steered people via its 'Your Voice, Your Choice' platform to an AI tool offering around 50 pre-defined responses, almost all opposing stricter EU tobacco rules; Pointer's Pangram analysis judged 35 per cent of consultation responses from fifteen countries AI-written. PMI says users kept full control and accuses anti-tobacco lobbyists of spreading fake news.
Education Minister Mihai Dimian said invigilators stopped several 2026 Bacalaureat cheating attempts, including with glasses fitted with video cameras and artificial intelligence. Meanwhile, Alba prosecutors ran 47 searches in five counties and detained five suspects they say ran a network since 2023 that photographed exam subjects, had them solved and dictated answers to paying candidates via hidden earpieces.
Fraudsters are using live deepfake video calls to impersonate Dubai's crown prince, Hamdan bin Mohammed, in romance scams, AFP reports, with researchers tracing some to Nigerian crime syndicates. One Filipino domestic worker lost a year's savings, paying 100,000 pesos for a supposed marriage certificate and 'royal membership card', before balking at a 60,000-peso hotel booking.
São Paulo's regional electoral court ordered, in a preliminary ruling, removal of a video depicting Governor Tarcísio de Freitas as the killer doll Chucky, citing signs of negative electoral propaganda and unlabelled AI use, in a complaint by Republicanos against PT state deputy Emídio de Souza and party bodies. The state PT says it never posted the video and calls it satire.
A Singaporean actress says a man used AI to create an image of the two of them at the beach with her in a bikini, then claimed she was his wife, posted it with a caption about raping her and said he wanted a protection order against her. Police confirmed she made a report; she criticised commenters who mocked her.
Bruno, of the sertanejo duo Bruno & Marrone, used Instagram on 2 July to deny rumours that he had terminal cancer with six months to live, spread through AI-generated montages showing Marrone and Bruno's son crying. 'It must be another Bruno,' he said, adding that he was very well and had a show in Mato Grosso the next day.
