AI for LinkedIn: My Exact Playbook

LinkedIn in 2026 is weird. The feed is flooded with AI-generated content, and somehow the algorithm still rewards it because engagement is engagement. But the people actually building real audiences and generating real leads are doing something different.

They’re using AI as a thinking tool, not a publishing tool. There’s a significant difference and it’s everything.

Let me walk you through what I do.

Why Most AI LinkedIn Content Dies in the Feed

You know the posts I’m talking about. The three-line-per-paragraph format. The “here are 5 things I learned from [famous person]” structure. The “let me tell you what X taught me about Y” takes. They’re everywhere. They perform okay because LinkedIn has a long tail of people who haven’t been on the platform enough to find them boring yet.

But they don’t build relationships. They don’t generate inbound leads. They don’t make someone want to follow you specifically because of your perspective. They’re content shaped like LinkedIn content.

Real LinkedIn influence comes from specificity, from genuine opinions, from posts that could only have come from you. AI can help you get there faster. It cannot get there without your input.

How I Use AI for Ideation (Not Writing)

Every week I do a 10-minute ideation session with Claude. Not to write posts - to find angles.

I describe what I’ve been working on, reading, thinking about, or arguing with someone about that week. Then I ask: “Based on this, what are 5 LinkedIn post angles that would be genuinely interesting to a senior marketing director? Each angle should be counterintuitive, specific, and take a clear position. Avoid common formats and obvious takes.”

The output is a list of angles. I pick one or two that resonate, discard the rest, and then write the posts myself - or use the angle as a starting point for a quick AI draft that I rewrite heavily.

The AI is finding the frame. I’m putting the substance in. That’s the right division of labor.

The Post Types That Actually Work for Me

Not all LinkedIn content is equal. I’ve tracked my posts obsessively enough to know what generates followers, what generates leads, and what generates likes from people who will never buy anything.

The “I was wrong” post. When I change my mind about something and explain why. These are my highest-engagement posts by a mile. People respect intellectual honesty, and it’s rare on LinkedIn. AI can help me structure the argument - here’s what I believed, here’s what changed, here’s the new position - but the opinion has to be mine.

The “here’s what nobody tells you about X” post. Works because it promises insider knowledge. AI helps me brainstorm what the non-obvious angles are. I validate from experience whether they’re actually true.

The specific result post. “We did X, here’s exactly what happened.” Numbers, context, what we tried, what didn’t work. These generate DMs from people who want to do the same thing. AI can help structure the narrative but the results are mine.

The contrarian hot take. Short, punchy, an opinion that a significant portion of LinkedIn professionals would disagree with. I use Claude to stress-test my hot takes before posting - “here’s my opinion, what’s the strongest counterargument?” That process either refines the take or kills it before it embarrasses me.

The AI Tools I’ve Tested for LinkedIn

Let me save you money. I’ve tried most of them.

Taplio: decent for scheduling and analytics. The AI writing features are average. I use it for the calendar view and the analytics, not the content generation.

AuthoredUp: good for formatting posts and A/B testing. The formatting tools are genuinely useful because LinkedIn’s native editor is terrible.

Taplio, Postwise, Lara, Hypefury’s LinkedIn features - they all have AI post generators. They all produce content that sounds like everyone else’s AI-generated content. I don’t use any of them for actual post writing.

What I actually use: Claude for ideation and drafting, AuthoredUp for formatting and scheduling, Shield Analytics for tracking what’s working.

The Commenting Strategy Nobody Talks About

This is where I think AI adds the most underrated value for LinkedIn.

Thoughtful comments on the right posts do more for LinkedIn growth than most people’s posting strategy. When you comment something genuinely insightful on a post from someone with a large following, you get exposed to their audience. If your profile is good, you convert that into followers.

I use Claude to help me write comments that are actually worth reading. I paste in a post, describe my genuine reaction or disagreement or additional experience, and ask Claude to help me articulate it clearly and concisely. The thought is mine. The articulation gets a little help.

This takes maybe 15 minutes a day. The compound effect over months is substantial.

Optimizing the Profile (This Part Is Unglamorous)

Nobody talks about this in the AI LinkedIn content because it’s boring, but it matters: your profile converts the interest your content generates.

I used Claude to rewrite my headline, about section, and featured section. I gave it my background, my goals, who I want to attract, and what I want them to do. I asked for 5 headline options that position me as a specific expert rather than a generic professional.

The test for a good LinkedIn headline: could this exact headline be used by 10,000 other people? If yes, it’s not specific enough.

I also asked Claude to identify the three things someone visiting my profile needs to believe about me before they’d send a connection request or DM. Then I checked whether my profile actually communicated those things. It didn’t, in a few places. I fixed them.

The Volume Question

How often should you post? This is the question I get most. Here’s my honest take.

Consistency beats frequency. Two posts per week that are genuinely interesting beat five posts per week where two are good and three are filler.

AI makes it tempting to post more because producing content is faster. Resist this. Quality caps how much your following trusts you. Quantity only helps if the quality is there first.

I post 3 times a week. Every post has a point of view. None of them are just “here are 5 tips.” That standard takes discipline when you have a tool that can produce 5 posts in 5 minutes.

The discipline is the competitive advantage.