Fixing the five most common PR and AI problems

Transcript:

Hullo. It's Craig AI here. Call me CraigAI.

In these videos we'll look at the most common issues PR professionals tell me about with regards to problems in using AI. Think of me as being your agony aunt - well, agony uncle - to use UK tabloid lingo. To get us started we'll have a bumper video with FIVE of the most common issues.

Number one. Not giving AI enough context.

AIs need briefing, just like a junior staffer or work you give to an agency. If you don't guide them, you get bland, generic output. Remember, AI pulls in an average of everything so if you want it to stand out, make it distinctive - and that means some personalisation of output.

Number two. Blind trust.

People copy and paste AI text without fact-checking. That's dangerous. It can make up data, quotes, even sources. Always check - or get the LLM to cite sources. Even then, check.

Number three. Losing the brand voice.

AI often defaults to a corporate, robotic style. If you don't edit or have great prompts, your message won't land.

Number four. Narrow use.

Too many stop at "write me a press release or social media post - or video script (ahem)." Come on people! It can do so much more than that! AI can test angles, predict journalist questions, even help with localisation. Don't just think of it as a remote typewriter.

Number five. Forgetting the risks - because there's plenty.

AI can plagiarise. It can breach confidentiality. It can damage reputation if you don't double-check. There's actually another outstanding fail that I see and hear of nearly every day, but we'll save that for the next video.