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
Why Judgment Matters More Than Ever in the Age of AI
AI has made it easier than ever to produce something that looks finished with clean sentences, a confident tone, and logical flow at first glance. When that work goes out the door, it doesn’t belong to the tool. It belongs to whoever approved it.
A new Harvard Business Review article examines how workers develop judgment in the AI era and finds a clear divide. It finds that people who know how to evaluate work tend to get more value from AI, while those who don’t struggle to tell whether what they’re seeing is actually good.
In communications, that gap shows up fast. AI can explain an industry or summarize a trend, but once it starts generating real work — pitches, contributed content, messaging — judgment becomes the deciding factor. A pitch can be clean, confident, but completely ineffective, and still get sent because nothing looks wrong. An argument can sound airtight while skipping a key assumption that makes the whole thing collapse under scrutiny.
This is where critical thinking matters. AI writes with authority even when it’s wrong. If someone rubber-stamps an output because it sounds polished, the tool actually puts their credibility at risk. Communicators need to dig beneath the words and make sure the intent is clear, that there aren’t important elements of the story missing, that the output would actually resonate with the intended audience.
We’re already seeing what happens when AI outputs go unquestioned. Judges have fined lawyers for submitting AI-generated filings that cited cases that never existed. The work looked legitimate. It wasn’t. And the damage was reputational.
This isn’t limited to edge cases. In his annual letter, YouTube’s CEO called out managing AI slop as a top priority. When platforms operating at that scale are worried about low-quality AI content, individual communicators should be too.
Experienced communicators can spot this immediately. The structure feels robotic. Ideas repeat. Storytelling never quite builds. Fixing that is about using judgment.
AI can help you get to a first draft faster. It cannot decide whether that draft deserves to exist, and it won’t warn you when it shouldn’t. That responsibility still sits with the person using it. The more AI becomes part of the workflow, the more valuable judgment becomes, and the harder it is to fake.
Elsewhere …
- The World’s First AI-Only Social Media is Seriously Weird. The Next May Be Even Stranger
- Anthropic Is About to Drop Sonnet 5 During Super Bowl Week
- OpenAI Gets a Meta-Style Makeover with Ads
Tips and Tricks
🌐 Did you really read that link?
What’s happening: As covered above, AI will confidently say it did something even if it didn’t. That can be particularly consequential if it’s lying about reading a link that includes useful context.
Keep an eye out: Both ChatGPT and Claude show (some of) their work as they consider a prompt and generate an output. While ChatGPT’s dot pulsates or Claude’s asterisk dances, you should see some text temporarily appear that says “Searching web” or “Fetching site.” If you don’t, be skeptical that it actually visited a link you shared.
Try this: Ask either platform explicitly whether it actually read a link, and it tends to give an honest answer. If it can’t read the page, you may need to upload it as a PDF file. If it can read the page, add to your custom instructions that anytime you share a link, it should visit that link rather than just guess the page's contents based on its URL.
Quote of the Week
“She’s learned to conceal her AI use rather than evaluate it. She’s developed shame instead of judgment. And when she graduates into a workplace where AI tools aren’t contraband but required, she won’t know how to think critically about their outputs. She’ll either avoid them entirely and fall behind, or use them uncritically and produce work she can’t defend. Neither option serves her well.”
— Rachel Toor, Contributing Editor at Inside Higher Ed and Professor of Creative Writing, in a column that includes how one of her students uses AI
How Was This Newsletter?
