Gennaro Brooks-Church Examines 5 Reasons Underwriters Turn to AI

00:00:00: Welcome to your custom deep dive.

00:00:02: Today, our mission is unpacking why the notoriously traditional world of insurance underwriting?

00:00:08: it's just rapidly adopting AI.

00:00:11: I mean you probably picture self-driving cars when you know, a mountain of scanned PDFs.

00:00:17: Yeah totally but underwriting is actually ground zero for this tech.

00:00:21: we're using insights today from Genaro Brooks Church.

00:00:24: he's a New York based executive leader and founder of Casimir in Brooks Energy

00:00:31: Right, and his analysis really focuses on this massive structural bottleneck.

00:00:36: Underwriters are just absolute buried in unstructured paperwork.

00:00:39: We're talking custom spreadsheets, messy emails poorly scanned documents.

00:00:42: Just the works.

00:00:43: Oh it's a total mess.

00:00:44: But AI can instantly scan an organize that entire unstructured avalanche into usable formats Which you know is a huge relief for these teams.

00:00:53: It's basically

00:00:54: having this tireless, super-powered assistant.

00:00:57: Like if you free an underwriter from manually copying and pasting data all day.

00:01:00: they can actually focus on their high value work.

00:01:03: Right things like building broker relationships or actually evaluating the complex accounts because evaluating risk requires following strict corporate standards.

00:01:12: but humans well we suffer from fatigue.

00:01:15: Oh for sure By the time you're reviewing your fiftieth messy PDF of day, You might miss a subtle detail or bend rule without meaning to.

00:01:24: Exactly!

00:01:25: And that is where the pivot-to consistency happens.

00:01:28: The AI acts as this steady co-pilot.

00:01:31: It doesn't just read the document, it maps the extracted data directly against a rigid algorithmic set of guidelines.

00:01:39: So its systematically flagging missing datapoints and highlighting unusual risk factors?

00:01:43: Yeah which guarantees absolute uniformity across millions.

00:01:49: If it's rigidly checking everything against standard rules, how does it handle the unwritten rules?

00:01:54: Like I'm thinking about those thirty-year veterans who know the highly specific risk factors that aren't spelled out in any company handbook.

00:02:01: Well that brings up a classic institutional knowledge strain you.

00:02:04: when senior staff retire they usually take decades of industry wisdom right out the door with them

00:02:09: which is huge problem for these older institutions.

00:02:12: It's massive but Brooks Church notes that modern AI actually captures those nuanced decision patterns over time.

00:02:21: By feeding the AI historical data, you're essentially mapping thousands of past approvals and denials.

00:02:28: Oh!

00:02:28: So it's learning specific variables that drove those past human decisions

00:02:33: Exactly.

00:02:34: The system builds a behavioral model in an institution.

00:02:37: best underwriters so new hires can tap into collective memory from day one.

00:02:42: I mean, i have a really hard time believing insurance companies are just handing the keys over to a predictive model though.

00:02:47: We hear so much about AI giving random answers or you know hallucinating

00:02:51: Right!

00:02:51: The Hallucination risk is real.

00:02:53: Yeah relying on an automated system for multi-million dollar policy evaluation sounds like massive liability.

00:02:59: How can they trust it?

00:03:00: Well They Can't Afford A Black Box Where Data Just Goes In And A Mysterious Answer pops out.

00:03:05: Rook's church emphasizes what he calls the glass box approach.

00:03:09: A glassbox, like total transparency?

00:03:12: Exactly!

00:03:12: It is the enterprise equivalent of a math teacher forcing his student to show their work in the margins.

00:03:17: Modern AI provides transparent verification

00:03:20: So it not just giving you final answer but actually has provide receipts.

00:03:24: Yes

00:03:24: every single extracted revenue figure or flagged liability risk links directly back to its original source document.

00:03:31: Oh wow Down to the exact page and

00:03:35: paragraph.

00:03:36: That transparent verification is the entire mechanism of trust.

00:03:41: The human underwriter can instantly trace the AI's logic back to raw data.

00:03:47: It proves its work, which keeps Human Judgment at the absolute center.

00:03:51: as a final checkpoint It really changes how you look at automation, like whether you are dealing with chaotic manual data entry in your own job or worrying about a senior colleague retiring and taking all their secrets to the grave with them.

00:04:07: Yeah this glass box approach shows how technology can actively preserve human wisdom rather than just replacing human workers.

00:04:14: it's a fundamental shift moving from AI as to AI as a permanent custodian of institutional memory.

00:04:23: Exactly, it preserves what makes the institution valuable in the first place?

00:04:27: It does leave you with a fascinating question to chew on though.

00:04:30: if an AI can perfectly map and preserve senior underwriters exact judgment patterns inside this transparent glass box What happens when the AI's collective institutional memory eventually outlives?