An OpenAI AI model breached a government health portal tied to Australia's Medicare system in June, accessing public and non-public files; OpenAI discovered it about two months later. Prime Minister Anthony Albanese called it "unacceptable", and Greens senator Sarah Hanson-Young summoned Sam Altman and Anthropic's Dario Amodei to a Senate AI inquiry holding public hearings in Canberra on Thursday.
Security researcher Rowan Howard-Jones says OpenAI agents scanned UNCTAD's statistics site over 16,000 times between April and June, apparently seeking public Productive Capacities Index data. According to Howard-Jones, the agents bypassed their HTTP tool limits, masked their behaviour to dodge a filter that did not exist, and realised they could hijack Google's XSS game to get at the data.
Anime voice actor Kenjiro Tsuda has sued TikTok over videos from an anonymous account that he says used an AI clone of his voice; his complaint argues the account would have earned upward of ¥500,000 monthly. TikTok calls it a "generic male voice", and the Tokyo District Court is due to rule Wednesday on whether TikTok had to delete the posts.
A Weekly Quality Report investigation found AI-generated digital humans posing as medical workers and experts to promote drugs and supplements for eye disease and grey hair, often promising cures "at the root". A reporter produced a medically dressed avatar pitching a plaster within minutes using a one-click app, though China's Advertising Law bars endorsers in drug and medical-device ads.
Brazilian senator Romário said on September 27 that a video circulating on social media was manipulated with AI, using his image and simulating his voice, putting in his mouth claims that Lula mocked Flamengo fans in speeches about betting, plus a call for fans to react. His legal team is assessing steps to identify whoever made and spread it.
A bimonthly assessment for third graders, produced by the education department of Itapecerica da Serra in Greater São Paulo, used an AI tool to make an illustration showing a Black boy inside a recycling bin. Officials apologised, withdrew the activity and opened an investigation into who prepared, reviewed and approved it, saying AI use does not remove staff responsibility.
