AI Scams Are Getting Smarter, and Healthcare AI Has an Outcome Problem

I’ve been watching the AI scam landscape evolve since ChatGPT first dropped, and honestly, it’s moved faster than I expected. What started as cybercriminals using LLMs to write slightly less obvious phishing emails has turned into a full-blown industrial operation. We’re talking automated vulnerability scans, hyperrealistic deepfakes, and phishing campaigns that adapt in real time.

Rhiannon Williams over at MIT Tech Review nails the current state: organizations are drowning. The volume of attacks is up, and AI makes them cheaper and faster to execute. Worse, the tools keep improving while more bad actors jump on board. If you’re in security, you already know this isn’t a future problem—it’s a today problem.

This is one of those rare cases where “AI arms race” isn’t hyperbole. Defenders are using AI too, but the asymmetry is brutal. Attackers only need one success; defenders need to block everything.

Meanwhile, on the healthcare side, Jessica Hamzelou raises a question that doesn’t get asked enough: Does AI actually help patients? Doctors are using AI for notetaking, scanning records, flagging at-risk patients, interpreting X-rays. Studies show these tools are accurate. But accuracy isn’t the same as better outcomes.

We don’t know if patients live longer, recover faster, or experience fewer complications because of these tools. That’s a gap that should worry everyone. It’s easy to deploy a tool that works well in a controlled study. It’s harder to measure whether it changes anything meaningful in a chaotic hospital setting.

I’ve seen this pattern before—new tech gets adopted because it’s shiny and seems useful, then years later we realize we never asked the hard questions. Healthcare AI deserves better.

Also worth noting: DeepSeek finally dropped V4. Preview versions are out, and they’re claiming it’s the most powerful open-source platform yet, rivaling closed models from OpenAI and DeepMind. It’s also adapted for Huawei chips, which is a geopolitical signal as much as a technical one. More countries are moving to restrict kids’ social media access—Norway is enforcing a ban, the Philippines might follow, and there’s growing pushback against AI in US schools.

Lots happening. The scams and healthcare stories are the ones I’d keep an eye on.