AI Exposure and the New Map of Work: What the Philadelphia Fed Data Reveals

I’ve been reflecting on two new studies that map not just what AI can do, but where it’s already reshaping the labor map.

A new Federal Reserve Bank of Philadelphia study reveals how deeply generative AI is reshaping work across our region, and what that means for the next generation.

A pattern emerges - exposure isn’t distributed evenly:
▪️The median occupation carries a 𝟬.𝟯𝟬𝟳 𝗲𝘅𝗽𝗼𝘀𝘂𝗿𝗲 𝘀𝗰𝗼𝗿𝗲, meaning about 𝟯𝟭 % 𝗼𝗳 𝘁𝗮𝘀𝗸𝘀 are now touchable by AI.
▪️Jobs requiring a 𝗯𝗮𝗰𝗵𝗲𝗹𝗼𝗿’𝘀 𝗱𝗲𝗴𝗿𝗲𝗲 are nearly 𝟯× 𝗺𝗼𝗿𝗲 𝗲𝘅𝗽𝗼𝘀𝗲𝗱 (≈ 45 %) than those without (≈ 14 %).
▪️𝗧𝗿𝗲𝗻𝘁𝗼𝗻–𝗣𝗿𝗶𝗻𝗰𝗲𝘁𝗼𝗻 and 𝗦𝘁𝗮𝘁𝗲 𝗖𝗼𝗹𝗹𝗲𝗴𝗲 lead in exposure; 𝗩𝗶𝗻𝗲𝗹𝗮𝗻𝗱 and 𝗚𝗲𝘁𝘁𝘆𝘀𝗯𝘂𝗿𝗴 trail.
▪️Exposure also varies across professional demographics, with 𝘄𝗼𝗺𝗲𝗻 𝗶𝗻 𝗰𝗲𝗿𝘁𝗮𝗶𝗻 𝗿𝗼𝗹𝗲𝘀 more represented in AI-affected work.

And it’s not just where we work - it’s who’s entering the workforce.

A separate Stanford Digital Economy Lab found early-career workers (ages 22–25) in these high-exposure fields saw a 𝟭𝟯 % 𝗿𝗲𝗹𝗮𝘁𝗶𝘃𝗲 𝗲𝗺𝗽𝗹𝗼𝘆𝗺𝗲𝗻𝘁 𝗱𝗲𝗰𝗹𝗶𝗻𝗲 since 2022, even as seasoned workers held steady.

But high exposure doesn’t always mean high risk.
It can signal early access to transformation: if regions and institutions act fast enough to turn automation into augmentation.

💡𝗥𝗲𝗮𝗱𝗶𝗻𝗲𝘀𝘀 𝗶𝘀 𝘁𝗵𝗲 𝗿𝗲𝗮𝗹 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁𝗶𝗮𝘁𝗼𝗿.
Regions investing in digital fluency now will convert exposure into advantage later.

“𝙷̶𝚘̶𝚠̶ 𝚖̶𝚊̶𝚗̶𝚢̶ 𝚓̶𝚘̶𝚋̶𝚜̶ 𝚊̶𝚛̶𝚎̶ 𝚊̶𝚝̶ 𝚛̶𝚒̶𝚜̶𝚔̶?"
→ “𝗛𝗼𝘄 𝘃𝗶𝘀𝗶𝗯𝗹𝗲, 𝗮𝗻𝗱 𝗹𝗲𝗴𝗶𝗯𝗹𝗲, 𝗮𝗿𝗲 𝗼𝘂𝗿 𝗽𝗲𝗼𝗽𝗹𝗲 𝘁𝗼 𝘁𝗵𝗲 𝘀𝘆𝘀𝘁𝗲𝗺𝘀 𝘀𝗵𝗮𝗽𝗶𝗻𝗴 𝗼𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝘆?”

Because this transition doesn’t have to divide.
It can connect - if we choose to build bridges between human potential and machine perception, between data and dignity.

Careers aren’t disappearing.
They’re being rewritten between the data points.
And every region deserves the chance to be seen - and supported - to get there.

-L. Fitzpatrick