Magistrate clears U.S. Bank's former AI chief to pursue discrimination and retaliation claims

Neither dismissal touches whether U.S. Bank did anything wrong. The defamation count went out on the calendar, having missed the statute of limitations ^[7], and the blacklisting count went out on a definition: Keesler found that when a U.S. Bank employee spoke about Srini Nallasivan to a firm on the verge of hiring him, the speech was not "unsolicited" and therefore fell outside North Carolina's blacklisting statute ^[8]. That is a genuinely interesting term of art, because a reference check is solicited by construction, which means the conduct alleged here (an offer with a generous pay package withdrawn four days after a U.S. Bank employee spoke negatively, per the complaint ^[3]) is reachable in North Carolina only inside the defamation window, and Nallasivan arrived outside it.

So the surviving claim is the cheaper-sounding one, or rather the one whose arithmetic is easier to state. He pleads lost wages and benefits plus emotional distress amounting to at least $100,000 ^[10]; his supervisor allegedly promised severance of more than $100,000 conditional on his helping with the transition ^[1]. The floor of the case and the sum the bank allegedly withdrew are the same number ^[20]. The acceleration compounds it: a January 2024 notice with a May last day, moved a month later to mid-March with no severance ^[1][2], takes roughly two months of pay out on top ^[18], which Nallasivan's lawyers call "a classic bait-and-switch" ^[2].

What makes "the role was eliminated" a shakier defence here is the sequence of events leading up to it, not the elimination itself. A September 14 letter accused him of being combative, aggressive, uncollaborative, of hiding information and circumventing regulations ^[16]; in November he learned another employee, a white man according to the complaint ^[17], would take over the chief of AI role he had held since 2019 ^[17]; the elimination notice came about four months after the letter ^[21]. Keesler's wording is recommendatory, that the claim "should, at this stage, be allowed to proceed" ^[5], and "not time-barred" is a finding about a clock rather than about conduct ^[5]. U.S. Bank declined to comment on the ruling ^[11] and in August 2025 said the job was eliminated after "a robust and considered organizational review" ^[12].

This could go three different ways from here. Maybe the recommendation is not adopted as written and the blacklisting count comes back. Maybe U.S. Bank settles before internal documents about the November succession decision are produced, which is what most defendants with a fourteen-year-veteran plaintiff (twelve years at Bank of America, two at AIG before his 2018 hire ^[13][19]) tend to do. Or maybe discovery shows the consolidation was approved before the supervisor was pushed out in September 2023 ^[15], in which case the chronology is coincidence.

This is probably wrong, but the useful read is a pricing one: eliminating an AI leadership post is a defence that costs a motion, and after Thursday it costs discovery. The complaint alleges glowing annual reviews through 2022 ^[14] and asserts the actions "were intended by U.S. Bank to replace Mr. Nallasivan, a person of color" ^[23], which is exactly the sort of contested narrative that motions cannot resolve. What would falsify it: a documented reorganisation dated before September 2023 in which the AI duties were folded into an existing role, at which point the successor's race and the letter's timing stop doing work.