If Wall Street Can Demand It From BigLaw, So Can You
Custodia Advisory | Private Equity | Legal & Fiduciary Spend
Goldman Sachs, Morgan Stanley, and Citi have told their outside law firms something that should sound familiar to anyone who's read our Knowledge posts: if the work is taking less time, the bill should be smaller. According to reporting from the Financial Times this week, all three banks are pushing elite law firms toward new commercial arrangements, arguing that AI has meaningfully cut the time needed for research, document review, and discovery — and that legal fees haven't caught up with that reality.
The specifics are worth sitting with. Citigroup's global head of legal, Adam Meshel, put the logic plainly: if AI has cut the hours worked on a matter, costs should fall "significantly" per transaction. Morgan Stanley's general counsel, Eric Grossman, went further, describing BigLaw's traditional compensation model as "extraordinarily unstable." Goldman Sachs, for its part, has reportedly started asking outside counsel directly how much AI is actually saving them on Goldman's own matters — and using the answer as a lever in fee negotiations.
The timing makes the standoff sharper
This is landing at a genuinely strange moment for legal billing. Associate hourly rates at major firms have climbed past $700 this year, up roughly a third since 2023, with partner rates not far behind. Automation is accelerating at the same time headline billing rates are hitting record highs — which is precisely the contradiction Wall Street's largest clients are now refusing to simply accept. The efficiency gain from AI is real and measurable. The question these banks are asking, bluntly, is who actually captures it: the firm that bought the technology, or the client who's still paying pre-automation rates for post-automation work.
Worth noting: this isn't a wholesale rejection of legal spend. Where the work is genuinely complex, technical, and requires real judgment — the kind of high-stakes transactional and advisory work AI can't meaningfully shortcut — these same banks remain willing payers. The pushback is targeted specifically at the leverage model: the routine, high-volume, junior-staffed work that AI now does in a fraction of the time, but that continued to get billed as though it still took a fraction of a career to master.
Why this matters beyond the banks
It's tempting to read this as a story specific to a handful of the world's largest financial institutions and their outside law firms. It isn't. The underlying principle applies with equal force to every platform paying for legal and fiduciary services, at any scale: a fee structure agreed years ago doesn't automatically stay fair just because nobody's revisited it. Efficiency gains — whether from AI, from process improvements, or simply from a provider handling a matter type for the hundredth time rather than the first — don't pass through to the client unless someone specifically insists that they do.
This is, at its core, the same argument behind proper OCG enforcement. Rate cards don't self-correct. Fee structures don't automatically reflect current reality. If Goldman Sachs, with all its leverage and legal spend, has to explicitly demand that efficiency gains get shared rather than quietly absorbed by the provider, it's a fairly clear signal that no platform — however sophisticated — can assume its own legal and fiduciary billing is keeping pace with what the work actually requires today, unless someone is checking.
The parallel on the fiduciary side
There's a version of this conversation worth having about fiduciary services too, and it's one we don't think enough platforms are asking. Much of AML/KYC administration, document preparation, and routine compliance monitoring is exactly the kind of process-driven, repeatable work that technology has made faster over the past several years. Yet fiduciary fee structures, in our experience, have shown far less willingness to reflect that efficiency than even BigLaw is now being forced to concede. If Wall Street's biggest banks can successfully press elite law firms on this point, there's no reason platforms of any size shouldn't be asking the same question of their fiduciary providers.
The lesson isn't really about AI. It's about who's willing to ask the question in the first place — and who has the standing, and the distance from the relationship, to actually press for an answer.
Custodia Advisory helps private equity and asset management platforms apply exactly this kind of scrutiny to legal and fiduciary spend. Enquire about a complimentary review of where your current fee structures may not reflect current efficiency.