Two Billion Dollars, Then Two Directors

On December 27, 2024, The Bank of Nova Scotia wired approximately $2.0 billion into KeyCorp and became the owner of roughly 14.9 percent of an American bank holding company with about $190 billion of assets, a thousand branches and customers in fifteen states. The Federal Reserve had approved the purchase fifteen days earlier.

The stock was not the only thing that changed hands. That same day KeyCorp's board went from fourteen directors to fifteen, and two new names took seats. Neither had been found by a search firm. Both were placed there under the designation rights Scotiabank had negotiated into the investment agreement announced the previous August.

One of them was a Scotiabank group head, which is what a strategic investor's seat usually looks like.

The other was Somesh Khanna, who had never worked at either bank.

The Man With No Bank

Khanna, then 60, had spent his working life on the outside of institutions like this one, being paid to explain them to themselves. The KeyCorp filing lays the record out with a lawyer's flatness. He co-led McKinsey & Company's Global Banking and Securities Practice from 2018 until 2023 - described in the same filing as the largest global practice at the firm. From 2014 to 2018 he ran McKinsey Digital for Financial Services worldwide. He led the New York office in 2011 and 2012, and the Property and Casualty Insurance Practice from 2009 to 2011. From 2007 to 2024 he was a client service team leader on banking, insurance and private equity accounts, which is the consulting industry's phrase for the person whose name is on the relationship.

Before McKinsey he had actually run something: a digital business at eCredit, selling online credit and collections tools, and working on trade receivables securitization in partnership with J.P. Morgan.

His degrees trace back through Kellogg at Northwestern, the Indian Institute of Management in Kolkata, and a bachelor's in chemical engineering from the Indian Institute of Technology in New Delhi.

There is one thing the record cannot settle. KeyCorp's filing states plainly that he joined McKinsey in 2000. The Securities and Exchange Commission exhibit LPL Financial filed thirteen months later states that his tenure there began in 1998. Both documents are signed. They cannot both be right. It is a small thing, and it is the only visible seam in an otherwise seamless record.

What He Left First

He had already stepped away before the KeyCorp seat arrived. In June 2024, six months before Scotiabank's money closed, Khanna became Co-Executive Chairman of Apexon, a technology services firm. The consulting partner became an operator, or at least the chairman of one.

Then the bank seat, and with it the Risk Committee - the committee that exists to say no.

Two Point Three Trillion Dollars

Thirteen months after Cleveland, San Diego. On January 8, 2026, LPL Financial Holdings elected him an independent director. This is a different kind of institution: not a bank but the plumbing beneath other people's advice, supporting more than 32,000 financial advisors and roughly 1,100 financial institutions, servicing and custodying approximately $2.3 trillion of brokerage and advisory assets on behalf of about eight million Americans.

His statement on joining was written in the register of the industry he had spent decades billing. "LPL plays a critical role in empowering financial professionals to help millions of Americans progress toward financial security." Then the sentence that is either boilerplate or belief, and reads more like the second: "It's a noble purpose and one that LPL has proven to lead across the industry."

By then he was also on the board of Persado, and, since February 2026, the board of an artificial-intelligence company called ASAPP.

The Line He Put Down

On June 30, 2026, ASAPP made him chairman. The release ran the usual furniture - a veteran leader, a next phase of growth - and then Khanna said something that companies do not normally let a new chairman say in their own press release.

"I've spent 30 years watching companies buy software that promised to change everything and changed very little."

Read that again with his resume next to it. For most of those thirty years he was on the selling side of that transaction, running the practice that told banks and insurers what to buy and how to install it. He is describing a failure rate, and he is describing it from inside.

He did not stop there. "The pattern is familiar: big announcements, difficult implementation, modest results."

Then the pivot every chairman has to make: ASAPP, he said, "breaks that pattern with real results, loyal customers, and a consistent roadmap." His job now is to prove that the exception is real, in a category he has just publicly told the market usually is not.

Three Seats, One Table

Look at the arrangement whole. He sits on the risk committee of a $190 billion bank. He sits on the board of the platform that custodies $2.3 trillion for eight million people. And he chairs an enterprise AI company whose natural customers are institutions that look exactly like the first two.

None of it is hidden; all three seats are disclosed, and independence at each is a matter for the respective boards. But the shape is unusual. Most people who spend a career advising financial institutions retire into one board seat and a title. He took the advisory relationship and split it into three positions on three different sides of the same table - buyer, distributor, vendor.

The chair he holds at KeyCorp exists because someone else bought 14.9 percent of the company and reserved the right to fill it. He was the name on the other side of that clause.

Now he has to be worth it, at a company that just heard him say most of what firms like his sold changed very little.