Each week we’ll gather headlines and tips to keep you current with how generative AI affects PR and the world at large. If you have ideas on how to improve the newsletter, let us know!

What You Should Know

Beware the ai detectors

Writing has always been a subjective craft. These days, there’s judgment and scrutiny of not just the words, but who (or what) is behind them. AI detectors have become gatekeepers in classrooms, boardrooms, and newsrooms. These tools promise to sniff out AI-written content with near-perfect precision, but are often more smoke than fire. Research has shown that AI detectors are “neither accurate nor reliable” and are biased against non-native English writers.

OpenAI shuttered its detection tool two years ago, admitting to low accuracy, but that hasn’t stopped AI detectors from proliferating. Or making big claims. The Federal Trade Commission recently called out one company, Workado, for claiming 98% accuracy when its actual performance hovered around 53%.

There’s an even bigger issue than these tools’ unreliability. They reinforce a lazy binary: Human writing = good, AI writing = bad. That antiquated mindset misses the point. Bad content is bad, regardless of its author. Communicators shouldn’t be shaping their work to sidestep a half-baked algorithm. They should be focused on telling compelling, credible stories that move people. If AI helps you get there, great. If it doesn’t, don’t use it. But let’s stop pretending the presence of AI is the problem — or that these detectors are the fix.

Elsewhere …

• PODCAST: AI Turns Investory Guesswork into Science
• How AI Is Changing Search and Advertising
• OpenAI Scuttles For-Profit Transformation
• Takeaways from the AI Marketing Summit
• 1,000 AI-Enhanced Works Now Protected by US Copyright Law
• Google Eyes AI Chatbots as New Ad Territory

🔁 Create your own feedback loop

What’s happening: Whether you wrote something with AI help or built it from scratch, you don’t have to stop once you’ve got a draft. You can use tools like ChatGPT or Claude to create a feedback loop and test your work before it goes out the door.

Why it matters: Sometimes you spend so much time creating something that your eyes gloss over mistakes. AI can help you spot weak transitions, bloated sentences, or sections that just don’t land. It can be your second set of eyes.

Try this: After you’ve got a working draft, paste it back in and ask for feedback. Sometimes models will grovel and only offer positive feedback (OpenAI recently updated its 4o model to fix the sycophancy). If that’s the case, follow up and ask for constructive criticism. You can even add in instructions like “pretend you’re a writing coach” or “read this like a skeptic.”

You don’t have to take all the advice to heart, of course, but creating a feedback loop can help you know when you’re actually done with your draft.

Quote of the Week

“We’ve spent the past two years building applications that know how to use your data to help you answer a question and then give you a text answer back. I think we’re going to spend the next two years building applications that perform part of the work for you.”

— Eric Boyd, Corporate Vice President for AI Platforms at Microsoft, in a Q&A with ZDNET about the future of custom AI

How Was This Newsletter?