The Good
OpenAI has indefinitely shelved plans for an erotic chatbot, citing concerns from staff and investors about the impact of sexualised AI content on society. One adviser warned the company could be building a “sexy suicide coach” – AI that pairs intimacy with vulnerability, with no clear safeguards. The fact that internal pressure killed this before launch is a rare win for conscience over commercialisation. It won't be the last time this question comes up – but this time, the right people said no.
Comment: OpenAI also just shut down Sora because they were burning millions daily on a tool that could generate convincing synthetic video at scale. The market for “make any video you want” turned out to be smaller than expected – but also because the social value of that content is depreciating fast. Read on for The Feature that touches more on this.

The Bad
AI data centre investment is holding up a large part of the US economy, accounting for 39% of recent GDP growth. The Iran war is now threatening that through energy disruptions, private credit wobbles, and oil-driven inflation all at once. Those shocks ripple out to Australians through rising costs (⛽️🥲) and superannuation exposure to global tech stocks. But the war has also exposed the core vulnerability: the global AI boom was built on cheap Middle Eastern energy, and that advantage is no longer certain.
Comment: This could push companies toward solar, wind and more efficient infrastructure – which is exactly where Australia has a real (still unrealised) advantage.

& The Ugly
A US jury found Meta and Google liable for designing deliberately addictive platforms accessible to children. The verdict went straight to design: infinite scroll, autoplay, and algorithms engineered to space out dopamine hits like a slot machine, targeted at children with underdeveloped brains.
Comment: This is ugly on many fronts. Children were knowingly harmed for profit and, as a society, we have to sit with that. For Big Tech, the legal reckoning is just beginning: thousands of similar lawsuits are likely to follow. In Australia, our eSafety Commissioner is enforcing the U16s social media ban Big Tech lobbied against. This verdict just gave us more ammo.

Something has changed in how we relate to information online. Most of us can feel it.
AI-generated images, videos and voices are now cheap, fast and convincing enough to fool anyone on a bad day. The result isn't a world full of people believing everything they see. It's a world where people have stopped trusting anything at all.

That loss of trust has a cost. Democracy doesn't run on perfect information – but it does need a shared baseline of reality. When that erodes, so does everything built on top of it.

So it’s probably good that I sat down with Callum Harvey for a chat. He’s currently doing his PhD at the Oxford Internet Institute researching cyber threat intelligence and AI policy. His prior roles span the Australian Department of Industry, Science and Resources, CyberCX, and the Harris Cyber Policy Initiative at the University of Chicago. We talked about where we are at with mis/disinformation and deepfakes: what's working, what isn't, and why the answer probably starts smaller and closer to home than we think.

JK: I want to start with the landscape. When you look at AI-generated misinformation and disinformation right now, what are we actually dealing with?

CH: The way you can look at this in a very hopeful sort of way is that, yes, there is a lot of mis and disinformation online, but we also don't have a particularly great read as to whether any of it's effective. There is no easy way to measure the effectiveness of any kind of influence operation. That goes back to the Cold War if not before. If you read Thomas Rid's book, Active Measures, both sides were doing leaflet drops over East or West Berlin. It was a spectacle, but we don't know if any of that worked. And the same is true today.

JK: So we don't actually know if mis/disinformation and deepfakes are working?

CH: We don't. And that's actually kind of good news, in a way. But the inverse of that is you can't measure the policy response either. So people don't need to roll over and give up on this – they definitely shouldn't – because people do believe information that runs counter to fact, and that sucks. We need to build up people's media literacy to counter that. A lot of it is in the eye of the beholder, which I find really fascinating. A lot of people in government talk about how difficult it is to counter disinformation emerging from Russia and China, but we have no metric for whether it's doing what it says on the tin.

Read the full interview here.

If you’ve made it this far, thanks for joining us again. We’d love to know your thoughts – please hit us up. Otherwise, Happy Easter and see you in the next drop.

Yours in humanity,