analysisIs the new Princess of Wales' new photo 'fake'? The question is more interesting than the answerABy Ange LavoipierreTopic:Information Technology IndustryMon 11 Mar 2024Monday 11 March 2024Mon 11 Mar 2024 at 6:09amMajor news and photography agencies have removed a photograph of the Princess of Wales over concerns the image has been "manipulated".(Instagram: The Prince and Princess of Wales)Welcome to your new weekly tech update, from national technology reporter Ange Lavoipierre, featuring the best, worst and strangest news in AI, platforms, innovation and the online world.Sometimes, the paranoid-sounding theoryturns out to be correct.On Sunday morning UK time, the Princess of Walesposteda Mothers Day photo with her three children and, within minutes, social media wasawash with suspicionit was "AI-generated".Where was her wedding ring? Have her legs shrunk? Are Louis' crossed fingers in fact a SIGNAL to the PUBLIC?WhY ArE tHE LEaVES sO GREEN?Twelve hours later, four of the largest news agencies had issued a "kill notice" on the image, asking publishers to take it down.AP, the first to call it, said "it appears that the source [Kensington Palace] has manipulated the image in a way that did not meet AP's photo standards."AP's kill notice.(AP)It wasn't legs, leaves, or a missing ring that did it in the end, but a patch by Princess Charlotte's sleeve, bearing the telltale signs of a (poorly) edited photo. Cue meltdown.Royal intrigue aside though, an edited royals photo is not something I would normally think too hard about — I'm the kind of journalist who had to Google the children's names.From a broader view though, it captures something important about the deepeningtrust recession of the 2020s, which now seems bottomless, thanks to AI.Catherine, Princess of Wales, has been the subject of a suite of conspiracy theories since late January, when the palace announced she washaving abdominal surgeryand would withdraw from public duties until March 31.In that context, and arguably even outside of it, there were always going to be accusations like this.Even if the photo hadn't been edited, people would have said it was, and many would have believed them.Loading Instagram contentOn one level, who can blame them? Take the AI-generated photo ofDonald Trump with black voters, or theAI-fuelled attemptsto discredit Alexei Navalny's widow.As of March 2024, the US-based misinformation monitorNewsGuard was tracking 739"unreliable AI-Generated news and information websites".It sounds bad. It is bad. But as a fraction of the information economy, the amount of AI-generated material is still minuscule, and the proportion that's nefarious is smaller still.Still, there's now enough of it in the wild that every image, video and link — regardless of its provenance — is capable of inspiring suspicion. Especially if we don't like what it has to say.If anything, AP's kill notice is evidence that, for now at least, the systems we rely on to tell the difference are working.Yes, in this instance, the paranoid theory turned out to be (probably) correct, at least on some level.But the scary part is whether we'll bother to check next time.Loading...OpenAI: 'Nothing to see here'I'm paraphrasing, but speaking of paranoia, why else would OpenAIrelease a long-awaited reportinto itsembarrassing board meltdownon a Friday?For three months, independent investigators from the law firm WilmerHale have been trying to find out whether the board was right to fire chief executive Sam Altman over concerns he was being deceptive, and rehire him several days later.OpenAI chief executive Sam Altman.(Reuters: Elizabeth Frantz)The verdict? The board "acted within its broad discretion". It also found Sam Altman's conduct "did not mandate removal."It might be all we ever know about the investigation itself, as OpenAI has chosen to release a short summary of its findings, instead of the full report. Regardless, Sam Altman is nowback on the board, along withthree new female appointees.If you missed it, don't feel bad — that might have been the point. OpenAI very much want this story behind them so they can get on with the job ofbuilding AI that's better than humans.TikTok is fighting backAlmost exactly a year ago, we were reporting on apossible US TikTok banthat never eventuated.Now, there's growing support for a bill in the US Congress that would effectively ban the short form video app. Will this time be any different? The likelihood is increasing, but even if the bill is passed, you can still expect TikTok to challenge it with more energy than acancelled carnivore influencer on steroids.How do we know? ByteDance, TikTok's parent company, has already sentpush notificationsto recruit individual users to call politicians and ask them not to support it.A new bill before the Congress would effectively ban TikTok in the US.(Reuters: Dado Ruvic)Finally, if it's all too much…Please enjoythis OnlyFans subredditwhich is exclusively about literal air conditioning. It's our inaugural subreddit of the week. Feel free to send us yours.