We’ve all had a play with AI — turning a loved one into a Sesame Street Muppet (just me?) or digitally reviving a rust-ravaged barn find into a concours queen. But does AI really belong at the start of a genuine car restoration?

Rust Doesn’t Follow Rules
AI is a powerful assistant, but restoration happens in the real world. Can it offer practical, usable advice for classic car projects? In my view: no.

Rust doesn’t read manuals. Cars don’t age consistently. Every vehicle corrodes in different places, shaped by countless variables. That’s where AI struggles — it can’t make allowances for reality. It can list the “top 10” rust removers in the UK, but it can’t feel thin metal, spot a bodge, smell trapped damp or judge how far the repair really needs to go.

Why Restoration Isn’t Repeatable
Anyone who’s done more than one project knows restoration is never a repeatable process. Every used car hides a surprise. AI thrives on patterns and consistency — classic cars offer neither.

• Two identical models can suffer completely different corrosion

• Coastal cars bear no resemblance to dry-stored examples

• Old repairs change everything

No algorithm can account for all that.

Where AI Can Help
AI does have a role:

• Researching known weak points on specific models

• Surfacing workshop manuals and technical data

• Helping newcomers avoid obvious early mistakes

It can speed up the research phase — but you still have to read, question and apply that information properly.

Experience Still Wins
Real-world experience is priceless. If it’s your first project, join a local club, go to shows, talk to enthusiasts and speak directly with manufacturers and suppliers. Use AI when it helps — just don’t rely on it.

Nothing beats Old Man Paul in his 80s who’s restored ten cars and seen every mistake twice. Go and find him.