In this episode, Katie Daley Infante, who leads enterprise sales and partnerships at PerformLine, sits down with Kuno Tucker, a chief compliance officer with years of frontline experience building compliance programs in financial services, to talk about the shift from fixed, rules-based pre-publication review to AI-driven contextual review.
In this episode, we discuss:
- The shift from deterministic, if-then rules to contextual AI review that reads for meaning and intent, not just exact keywords, and where that shift delivers the most immediate impact for compliance teams.
- What to evaluate in an enterprise-grade AI vendor: data residency, explainability, auditability, and how often the vendor tests and corrects its models.
- The build-vs-buy decision, and a practical rollout path, internal testing, a small pilot group, then phased tranches—for getting advisor and compliance team buy-in.
Show Notes
- Learn more about PerformLine’s Pre-Publication Scanner: PerformLine Pre-Publication Scanner
- Connect with Ashley Cianci on LinkedIn: https://www.linkedin.com/in/ashley-cianci/
- Connect with Katie Daley Infante on LinkedIn: https://www.linkedin.com/in/daleykatherine/
- Connect with Kuno Tucker: https://www.linkedin.com/in/kuno-tucker-ma-llm-icd-6035355/
- Subscribe to PerformLine to stay connected to resources and updates: https://lp.performline.com/subscribe-to-performline
Subscribe to COMPLY: The Marketing Compliance Podcast
About COMPLY: The Marketing Compliance Podcast
The state of marketing compliance and regulation is evolving faster than ever. On the COMPLY Podcast, we sit down with the biggest names in marketing, compliance, regulations, and innovation as they share their playbooks to help you take your compliance practice to the next level.
Episode Transcript
Ashley Cianci:
Hey there, COMPLY Podcast listeners, and welcome to this week’s episode. In today’s episode, we’re sharing a conversation on the next era of pre-publication review, and the industry’s shift from fixed, rules-based review toward AI-driven contextual review that reads for meaning and intent, not just exact keywords. Katie Daley Infante, who leads enterprise sales and partnerships here at PerformLine, sits down with Kuno Tucker, a seasoned chief compliance officer with years of frontline experience building and running compliance programs in financial services.
Together, they dig into where contextual review delivers immediate impact, what to look for in an enterprise-grade AI vendor, the build-versus-buy decision, and how to roll out AI review in a way your advisors and compliance teams will actually adopt. As always, we appreciate you listening, and we hope you enjoy.
Ashley Cianci:
Afternoon, everyone, and thank you so much for joining us today. I’m Ashley Cianci with PerformLine, and I’m really glad you’re here. Pre-publication review isn’t new. Most compliance teams have been doing it in some form for years. What’s changing, and changing fast, is how it actually works. For a long time, reviewing marketing content before it went out meant running it against fixed if-then rules: spelling out every term and scenario in advance.
Over the last couple of years, the institutions we work with have pushed hard in a different direction: toward AI and contextual review that reads for meaning and intent, not just exact words. That shift is what today’s conversation is all about. We wanted to have this discussion out in the open, so I’ve brought together two people who see it from very different seats. Katie Daley Infante leads enterprise sales and partnerships here at PerformLine, and she spends her days with banks and financial institutions navigating exactly this transition. Joining her is Kuno Tucker, a seasoned chief compliance officer with years of frontline experience building and running compliance programs in financial services.
I want to be upfront about one thing: this isn’t a sales session. It’s a real, candid conversation between a technology partner and a trusted compliance leader about where pre-publication review is headed, what’s working, what to watch out for, and how to do this thoughtfully. PerformLine has been doing pre-pub review for years, so we care a lot about getting it right.
So with that, I’ll step back and hand it over to Katie and Kuno to get us started. Katie, over to you.
Katie Daley Infante:
Thank you so much, Ashley. We’re so excited to be here today. I want to set the context for everyone and talk about the elephant in the room: the monetary cost of getting things wrong, and why pre-pub review is necessary in this space.
We all know fines get the headlines, but they rarely start with a dramatic violation. Our own benchmark research with leading banks found that 65% of marketing compliance issues fall into just six categories: misrepresented free offers, bait-and-switch tactics, unsubstantiated claims, deceptive promotions, outdated APRs, and deceptive guarantee language. These aren’t large, exotic violations. They’re the routine output of marketing teams moving fast. They’re also exactly what pre-publication review was built to catch.
We’ve been doing pre-pub for years, but over the last 24 months, we’ve seen tier-one banks push hard for AI and move away from traditional, deterministic if-then-else rules toward contextual logic. I’m excited to unpack this: what should tier-one institutions actually think about when they implement AI at an enterprise level, given real GRC and governance constraints?
So Kuno, let’s kick it off. When you first started hearing “AI” and “compliance review” in the same sentence, what was your gut reaction?
Kuno Tucker:
Thanks, Katie, and it’s wonderful to be here. My first reaction to AI itself (not marketing yet, just AI broadly) was actually fear. I teach corporate governance at a university in Canada, and I was concerned about what people would do with AI without proper guardrails. Some of those fears have come true.
Fast forward, and people started talking about using AI for trading, portfolio management, and now marketing reviews. And I’m all for it. At a previous firm, when I joined, there was a lot of frustration among advisors because of slow response times and SLAs: you’d send an email in, wait for a reviewer to respond, send it back, it would go to another person, and so on.
Katie Daley Infante:
Yeah.
Kuno Tucker:
We adopted some other technology, not AI, which helped a little. So to answer your question directly: when I hear about AI in marketing reviews, I’m excited. I think it’ll be a real game-changer for a lot of firms, not just in efficiency and effectiveness (which matter) and compliance, but also in the advisor experience. So I’m excited.
Katie Daley Infante:
I think you touched on so many good points there. One in particular: that back-and-forth, the very human experience of needing everything urgently, and the question of how we review faster and more compliantly at scale. Compliance teams were sized for a world where marketing produced content at a human pace. That world has shifted dramatically. AI lets a single team generate far more copy, ads, pitch decks, and disclosures in less time than any review process was built to handle. And that volume keeps growing as AI gets adopted at the marketing level, the advisor level, you name it.
So how do you talk about this shift with your compliance teams, and now with your students, to prepare them for an AI-enabled world?
Kuno Tucker:
Yeah, they’re two different conversations. With students, it’s: embrace AI, but use it thoughtfully. There are numerous cases of lawyers using AI without reviewing the material themselves. A few have been disbarred, suspended, or sanctioned. So don’t use AI on its own; there needs to be a human in the loop.
With compliance officers and marketing specialists, one benefit I didn’t mention before is consistency. One complaint I heard often from advisors was inconsistent responses. Everyone’s had the experience of calling a call center, getting answer A, calling back five minutes later, and getting answer B. Advisors need not just quick turnaround, but accuracy and confidence that they’re compliant. AI enables efficiency and effectiveness, but also consistency and accuracy.
So for compliance specialists, the more tedious, mundane reviews can now be handled by AI, freeing people up for the more complex, higher-value work, which is also what people want to spend their time on. The AI handles the routine work in the background. I think the overall experience improves for the advisor, the compliance officer, and the firm. A triple win.
Katie Daley Infante:
You’re giving us so much gold here. I want to unpack that consistency point a bit more. Consistency matters not just for the end consumer’s experience, but internally too. You mentioned advisors having inconsistent experiences. The same is true for compliance reviewers. The way I’d review a 50-page pitch deck is probably different after my first cup of coffee than it is at 4:30, racing the clock with ten other things on my mind. AI does this at scale: it doesn’t get tired, it doesn’t take a sick day, and it isn’t juggling competing priorities. That consistency matters, especially as we push these reviews down to the content creators themselves.
That connects to a real shift we’re seeing: pre-pub review used to sit entirely with compliance. In the early years, the goal was to free up reviewers to focus on high-quality work, taking away the repetitive tasks. Now we’re seeing marketing content creators and advisor teams want to own that first line, since that’s where content is actually produced. In theory, if we can help them create better, more compliant content the first time, it reduces the number of revisions.
Have you seen this ownership shift moving downstream, away from compliance as the sole first line of defense?
Kuno Tucker:
Yes, Katie. It can get complicated, especially at more independent firms, but it happens at larger institutions too. The content creator might be an advisor with a strong point of view that may or may not be correct, and it may not be on-brand for the firm. That can get complicated, to put it mildly.
Having AI help with consistency, not just accuracy, can guide advisors toward brand-consistent language. Think about how much time marketing teams spend building a strong brand; everyone wants to be behind a strong brand. If you have what I’ll call a “rogue advisor”; and I don’t mean that dismissively, just someone who isn’t quite on-brand, that can create tension within the firm.
So from a brand perspective, AI can help create consistency of tone so the firm’s voice comes through as intended, while still letting the advisor get their own thoughts across.
Katie Daley Infante:
That’s right. Using AI to enable creativity, not hinder it. Compliance teams often get a reputation for saying no a lot, for boxing people in. That can feel constraining for a creative person. When AI is built into the creation process, it lets you iterate and create multiple versions of what you’re trying to say. The automated review can tell you in real time: yes, that’s compliant and on-brand, or, it’s slightly off, but here’s a generated suggestion that is, and you’re free to use it. It levels the playing field for everyone.
Kuno Tucker:
Exactly. There are obviously some keywords you can’t say, you can’t guarantee returns, for example, but there are other words that aren’t prohibited but don’t fit a firm’s brand culture. AI can guide advisors toward better language and more consistent brand messaging overall.
Katie Daley Infante:
Let’s talk about actually enabling that shift. We mentioned the push, from our clients and internally, away from deterministic logic, toward how we score for brand and compliance risk before assets go live.
The first move is away from deterministic logic and toward context intelligence. For anyone unfamiliar with the term: instead of a checklist; “if X is mentioned, then Y risk warning must appear within 50 words,” or “if X APR is present, Y disclosure must be in close proximity”, which is black-and-white and something deterministic logic handles well, we’re moving to AI, LLMs, generative AI, and image classification to look at compliance in new ways.
First, using LLMs to understand content: are we talking about this the right way, and does it match our compliance guidelines at scale? Instead of reviewing line by line, AI can triage all of it at once. Second, using image classification to look at logo, font, prominence, and color. It’s not enough that a disclosure or risk warning is present; is it legible, and is it easily understood by the person reading it?
Given all that, where do you see the most immediate impact from context intelligence at an enterprise institution?
Kuno Tucker:
That’s thoughtful, Katie. I know of a firm, no names, that received a six-figure sanction because the font on its disclosures was too small. People used to joke about that: what’s the “right” size? If you look at securities regulations in the U.S., Canada, or Europe, size isn’t specified. It just has to be legible. Legible to whom?
That goes back to accuracy and consistency. Deterministic logic on font size, at least, gives you certainty, no guessing, and that helps firms avoid avoidable fines and saves a lot of time. You’re not sitting there measuring font sizes by hand; the AI does that work, so the compliance officer can focus on higher-value work and trust the output.
Katie Daley Infante:
That’s a perfect example, and one we discuss often in demos. A clear, legible font size seems like something deterministic logic should handle: can you read it, yes or no? But AI unlocks much finer detail: comparing a slide’s 54-point header to 14-point body copy, and a 6-point disclosure that’s nowhere close to legible relative to everything else on the page. That lets us move away from a blunt rule like “must be at least 8-point, black text on white” and gives content creators more flexibility to produce content that looks a bit nicer than what we’ve traditionally seen.
What about reputational risk (political commentary, things that seem to require human nuance)? Where do you see AI helping there?
Kuno Tucker:
Yes, that comes up quite a bit. At every firm, you’ll have certain individuals who are opinionated, and they may even be correct, but it’s not something the firm wants in print. Maybe 90% of clients agree with a given view, but you’ll still have the other 10%, and I’m making those numbers up. Most firms want to stay apolitical and focus on the client’s best interest: we’re talking about investing, creating good outcomes, meeting financial goals. This isn’t a forum for political discussion, but some people can’t help themselves.
Having AI catch that and either filter it or reword it more palatably helps maintain the brand’s tone and avoid reputational risk, because once something’s on the internet, you can’t take it back. It’s best to catch it up front. A human reviewer might not catch every instance; AI will.
Katie Daley Infante:
Once it’s out in the wild, we could unpack that for another hour, but you hit the nail on the head. In your experience, if AI had been in place at a prior firm, do you think the first line of defense would have been stronger? Would it have caught things earlier, at the ideation stage, before a blog post or newsletter even came down for review?
Kuno Tucker:
You’re not going to stop someone from having their own thoughts, and that’s a good thing. But you can guide them before something off-brand or reputationally risky goes to print.
One other thing to remember: these are human beings. At one firm where I was CCO, I had advisors, some who didn’t even report to me directly, complain about specific compliance reviewers: “Person X doesn’t like my writing, I’d rather have Person Y review it. How about an AI instead?” That way, you always get the same consistent response, and no one can argue it’s a personal disagreement between them and a specific reviewer.
I think AI removes that argument and helps ensure the advisor’s output stays consistent, without advisors feeling like an individual reviewer is standing in their way, because it’s really the firm’s ethos, not a person’s opinion. Most advisors agree, when they sign on, not to write anything detrimental to the firm’s brand. But ten, fifteen, twenty years in, views can shift, or people get comfortable. Either way, the brand still needs protecting, and AI can help a lot there.
Katie Daley Infante:
You made a great point about people cherry-picking reviewers. We hear this often when teams are deciding whether to implement AI for pre-pub review. Larger institutions have multiple reviewers with competing opinions, and there’s human bias in review that people don’t always recognize. Taking out that person-to-person handover, and having an AI engine apply consistent standards and approved language, actually improves the culture. There’s no more middleman. If you disagree with a decision, you’re disagreeing with the guardrail itself, not a specific person’s judgment call.
Kuno Tucker:
Solid points. In my experience at other firms, the person straying too far off-brand usually isn’t wanted by other advisors either, because it’s their brand too, not just the firm’s. As an advisor, you want to be attached to a good brand, and you want your colleagues on-brand as well. It really protects everyone.
Katie Daley Infante:
We’ve talked a lot about enhancing compliance review and giving content creators tools to produce compliantly at scale, building that automated first line of defense and taking work off the manual review team. Let’s talk about the practical payoff: more accurate findings, fewer false positives compared to deterministic logic, and coverage of risk patterns that traditional checklists weren’t built to catch.
How much of your team’s time historically went to chasing false positives instead of focusing on real risk to the firm?
Kuno Tucker:
At my previous firms, we didn’t have AI in our marketing reviews. The technology was less mature then, so I don’t have that specific data point. But even without it: when an advisor said “I’d rather have Person X review my work than Person Y,” it was often because Person X understood context better and grasped the bigger picture faster. AI tools that do that well can go a long way, not just for speed and accuracy, but for reducing advisor frustration so they can focus on delivering the right marketing materials to clients on time.
Katie Daley Infante:
I want to shift into trustworthy, enterprise-built AI versus some of the newer startups on the market. In your opinion, what separates vendors when it comes to automated AI guardrails and pre-pub review? What’s paramount for an enterprise institution evaluating these technologies?
Kuno Tucker:
Setting AI aside, there are baseline things you want to think about with any technology vendor. I’ve been a U.S. and Canadian CCO, always domiciled in Canada, and in Canada, as in Europe and Asia, I imagine, we think a lot about data domicile. Even without exposing personal information, that matters.
For AI specifically, you want to know: how was the model built? What’s the underlying platform? How often do they test and correct it, and how do they respond when an issue is flagged? I think PerformLine does that well, but it’s something every firm should evaluate in any vendor, AI or not.
AI itself has progressed incredibly fast. From early LLMs to what we have today, it’s almost as if some advanced systems have something like consciousness. That said, AI can go wrong without a good vendor behind it, reviewing and testing properly. So I’d want any vendor to demonstrate they’re implementing responsibly and testing regularly, because a solution that’s solid today can go off the rails by version four, five, or six if testing lapses.
Katie Daley Infante:
I love this, because it goes down two paths. First, and I’ll admit this is a natural plug for PerformLine, we believe in enterprise-grade AI. We don’t want to slap on the AI buzzword for its own sake; things have to be tested and well-understood. In our case, that’s 18 years of experience feeding the intelligence behind our pre-publication review. No black-box models: we explain every decision. Why was this flagged as a violation? What regulatory statute does it relate to? What’s the correct language to use instead? And auditability: being able to defend that decision if an examiner asks why something was approved and who approved it.
The second path is build versus buy. We’ve seen a number of enterprises decide, “We have an AI team, let’s build it ourselves.” What they don’t always realize is that they’re now dependent on that team’s capacity to keep building V2, V3, V4, and to match the explainability a purpose-built vendor already has.
Have you seen that build-versus-buy argument play out in real time, whether in your courses or in practice?
Kuno Tucker:
100%, Katie. And I’ll say this with a smile: I fully understand that when I walk into a CCO role, I may not always be priority number one. I remember joining a firm that was fantastic from a compliance standpoint, and I’d worked with a number of their IT folks before. It was actually a trading algorithm firm, building trading algorithms for institutions.
I asked a couple of people I’d worked with before: why don’t we build more robust compliance tools ourselves? I can map it out. They smiled and said, “We could, Kuno, but look at the priority list for all our institutional clients and the algorithms we need to build for them. You won’t be at the top of that list. And even if we build it, can we keep up when there’s a rule change, a marketplace change, a new version needed? We won’t be able to service it the way a vendor whose core competency this is can.” In the end, we went to a third-party vendor.
There’s also speed to market: it’s not built overnight. You have to plan it, ideate it, build it, test it, deploy it, and then keep fixing it. That can take six months. A vendor that’s already built it should be closer to plug-and-play, and when a new rule set comes out or you want to change something, it’s a much smaller lift, with minimal disruption to the firm.
Katie Daley Infante:
Lots to unpack there. Let’s say you’re advising a firm (or a CCO) starting this transition. What’s the first signal that they’re ready for AI and contextual review? And as a follow-up, what’s the first mistake you’d tell them to avoid?
Kuno Tucker:
Great question. At a previous firm, if an AI solution had existed at the time, the signals would have been glaring. Advisors were unhappy with the marketing review process: the inconsistency, the SLAs, the turnaround times. We needed better technology, but AI wasn’t really an option yet, so we went with a non-AI third-party solution instead, which helped in a lot of ways. An AI-enabled solution would have gone even further.
The signals: end users (compliance officers) frustrated because their capacity doesn’t allow for quick turnaround, even though they’re not sitting idle; they genuinely want to get the work done, it’s a capacity issue. SLAs not being met, or not matching what advisors and analysts actually need. And it’s not just the reviewers: advisors and analysts need fast turnaround and aren’t getting it. In some cases, someone will tell you, as happened to me: “We’re just not going to get this pre-approved, we need to get it out.” That exposes the firm to real risk, and it’s why I moved quickly to find a better solution.
Other signals: a poor audit or exam result, or you can simply see that the consistency and accuracy of responses isn’t where it needs to be. Any one of these should prompt you to ask whether there’s a better way, and an AI-enabled solution can be that better way.
Once you’ve decided to move forward, the earlier discussion applies: don’t pick a vendor that hasn’t been in the space long, doesn’t test properly, or whose data residency you’re unsure of. You want a partner that keeps systems current and gives advisors and compliance users confidence in it.
One more consideration: growth. If a firm grows its assets under management, clients, or advisors, does that always mean hiring an eighth or ninth compliance officer? Or does it mean looking at technology that makes existing work more efficient so the team can absorb the extra load without adding headcount? This also comes up during mergers and acquisitions (I’ve been through a few), where different teams may have different views on consistency, accuracy, and SLAs. An AI solution can smooth that out, for both organic growth and M&A integration.
Katie Daley Infante:
Let’s say you’re going through one of those growth stages or an M&A integration. What’s the best path to deployment and user adoption? How would you roll this out?
Kuno Tucker:
Great question, because even the best solution can die on the vine if it’s not implemented thoughtfully. First: communicate, communicate, communicate. There should be no surprises. When you’re deep in a new vendor rollout, it becomes second nature to you, but for compliance officers, advisors, and analysts, it’s all new. You need to explain, early and often, why you’re implementing it and why it’s good not just for the firm and compliance, but for them: the advisor, the analyst. Most people don’t like change, so that context matters.
Next, make sure it’s the right tool and it’s been tested properly. Test internally with your compliance team and one or two technology-forward, positive “friendly advisors” who will actually use it and give constructive feedback. Once that’s done, pick a pilot group (I like somewhere between 10 and 20 advisors, depending on firm size) and run a pilot for 30 to 60 days with a clear feedback channel so issues surface quickly and the next iteration is well-tested.
Then, when you communicate to the broader advisor team, tell them what you did: we took on this tool to help you, we tested it internally, we tested it with a pilot group of your peers, and they loved it. Get testimonials if you can. Then launch thoughtfully: if you have 3,000 advisors, you probably don’t roll it out to everyone at once; you do it in tranches. Once you’ve gotten through a couple of tranches without issues and you’re getting good feedback, you can roll it out to everyone.
Katie Daley Infante:
You said so many good things, not just for the firm adopting this, but for keeping the vendor accountable too. Driving user adoption starts with everyone aligned on the same goal: why are we implementing this, what pain are we solving? If people aren’t bought in from the start, you end up with shelfware, nobody wants to use something if there’s too much change and no shared belief in the direction.
I also love getting that “squeeze” on the platform early: sending critical feedback back to the vendor. That builds internal champions, too. When the project manager isn’t in the room, you need end users who can explain why the team adopted this technology, and who can say, “We’re in this together, I can help you, you don’t have to go to the help desk every time.” It matters for implementation, and it’s great for us as the vendor to hear, we’re all aligned. It’s a partnership.
Kuno Tucker:
Completely agree.
Katie Daley Infante:
In line with that partnership, what other expectations do you have of your vendor? What matters most when evaluating a new technology provider?
Kuno Tucker:
There’s a lot there. As a chief compliance officer, when I want to bring in a new tool, I need budget approval, and that’s easier at some firms than others. You have to show it’ll be better for the firm overall and for the compliance department, but more importantly, that the front office (sales, analysts, advisors) will be better off, and by how much.
Having use cases and data to show the board or executive team matters, because you’re competing with other priorities. Someone might say, “That looks great, Kuno, but we’ve got three other things to focus on, let’s revisit this in two years.” Sometimes you have to say, “I don’t need it tomorrow, but two years is too long, can we talk about six to twelve months? Do we have the budget?”
And you need to prove the ROI: will this make us a better firm for existing advisors, help us recruit new ones, strengthen our reputation, support growth, and make M&A integration smoother? Those are the key selling points that matter to someone in my seat.
Katie Daley Infante:
That’s something we love helping with: building the business case. We’ve seen this technology implemented across hundreds of institutions, so we often have a bird’s-eye view of ROI that firms might not see on their own from a first-line rollout, which is useful when you’re building that case for budget approval. That’s the value of working with a trusted enterprise partner who’s done this many times before.
Kuno Tucker:
Exactly.
Ashley Cianci:
All right, thank you both so much. That was a fantastic conversation. If I had to sum it up in one line: this shift isn’t about AI reading faster. It’s about contextual judgment at scale, still grounded in your institution’s own knowledge, still explainable, and still something you can stand behind and defend to a regulator later.
Let’s close out with some final thoughts. Kuno, I’ll start with you.
Kuno Tucker:
I’m not usually one to quote movies, but in the movie F1, Brad Pitt’s character says in the pit stop: “Slow is smooth, smooth is fast.” I think that’s the key here: if AI makes the marketing process more frictionless and smoother, it actually makes it faster. So I don’t think you want to think only about speed and efficiency; think about effectiveness and the smoothness of the operation.
Katie Daley Infante:
I get so many bonus points for this because you didn’t even know I’m a huge F1 fan, that was the perfect quote to close on. It’s not always about being first or fastest. Controlled, expected results and repeatable scale, that’s what actually makes it fast.
Kuno Tucker:
Perfect.
Ashley Cianci:
Amazing, that’s a great note to end on. Thank you again to Katie and Kuno for such an open and practical conversation.
Thanks for listening to this week’s episode of the COMPLY Podcast. As always, for the latest content on all things marketing compliance, head to performline.com/resources. And for the most up-to-date industry news, events, and content, follow PerformLine on LinkedIn. Thanks again for listening, and we’ll see you next time.
