Punks & Pinstripes, with LeadGenius analyzed over 5,000 #AI people working at JPMorgan Chase & Co. Morgan Stanley, Goldman Sachs, Bank of America, Citi, and Wells Fargo, to determine who’s winning the war for AI Talent on Wall St? We wanted to know who's hiring the most, losing the most, and where this talent goes after they leave. Here’s what we learned.*** (Please read the methodology notes at the end to understand how we did this.)

1. No, really, read the methodology notes first, then come back and read our conclusions.

2. Wall Street Is Losing As Much AI Talent As It Hires. The typical bank loses four AI people for every five that it hires. This attrition is most pronounced at Goldman Sachs, which has hired 43 people and lost 75. In aggregate, over the past 24 months the biggest banks have hired 767 AI people and lost 708. JP Morgan is a significant outlier according to this analysis, hiring 318 AI people and losing only 204. 

3. Wall Street's AI Talent Would Rather Work For Amazon (or Google, or Microsoft). Amazon has lured more AI talent away from Wall Street than anyone else, hiring 153 people. Google comes second with 121, followed by Microsoft with 94, according to this data. If this pattern holds, Wall Street would lose ALL its AI talent to Big Tech in about four years.

1. JP Morgan is the Undisputed AI Talent Winner. By every measure, JP Morgan is the top destination for AI talent on Wall Street, according to this data. JPM has 1260 AI people on staff, at every level of seniority, more than twice Citi, who comes in second with 576. Equally, JPM is the only bank which has attracted more AI talent than it has lost, with a 3:2 hire:departure ratio.

5. If the current pattern holds Goldman will lose ALL its AI talent by 2026. According to this data Goldman has 182 AI people on staff and loses 35 more than it hires every 6 months. If this pattern persists Goldman will have depleted its AI talent pool in about 30 months.

6. Could Amazon build a challenger investment bank? If AI emerges as a significant competitive advantage in investment banking, then it’s not inconceivable that Amazon, Google or Microsoft creates a challenger investment bank. The three tech giants are the top destinations for AI talent leaving Wall Street. Amazon, in particular, has some history entering highly regulated new markets like healthcare. And Jeff Bezos began his career on Wall Street, at hedge fund DE Shaw.

Why is this happening? I'll be talking about this a lot more. But here's why I think Wall Street is struggling to hold on to AI talent. Being a technologist on Wall Street often means that you're an order taker, not an order maker. If you're a senior AI executive at Amazon you have pronounced say on the long term strategy of the company. On Wall Street, AI talent has historically implemented the executive team's strategy without determining what it ought to be.

Here's what needs to happen if you're a bank. My guess is that JP Morgan is benefiting from what I call the "Marty Chavez effect". Marty was a complete cultural outlier on Wall St, a Silicon Valley technologist who wasn't sequestered into a tech role. He eventually became the CFO of Goldman's powerhouse equities division, and above all was a magnet for Wall Street tech talent, who loved the flat, fast, innovation-forward culture that he nurtured. (He was also the most senior openly gay, Latino exec on Wall St.)

JP Morgan's Lori Beer and Oron Gill Haus have a reputation for being the technologists you want to work for if you're going to work on Wall Street. Everyone I know who's worked with them points to the same flat, innovation-forward, roll-up-your sleeves culture that Marty nurtured before he retired in 2019. The lesson is that cultural transformation is a critical pre-condition for digital (and especially) AI transformation.

Like this article? Please subscribe to my newsletter.

And, if you're an executive rebel learn more about my vetted community, Punks & Pinstripes here:

 ***Methodology Notes - To build this dataset, Lead Genius compiled data about hundreds of thousands of employees at the major banks and screened for AI engineers, product managers and executives. Sources included LinkedIn, and other news and social media posts announcing hires and departures. Like all unstructured data, this data is hard to compile and requires judgment. For example, we decided that if we could not categorize the level of seniority of an AI employee, we wouldn't include them in the analysis. This removed people who were consultants, interns, or whose seniority/ status was unclear. This made the conclusions more accurate, but also trimmed the sample size by about 65%. Also important: these banks are HUGE. The data is a small sample of what is likely a much larger number. Just being fully transparent.