#96 - The 4 steps to roll out AI

Sep 12

"We want to empower our own employees to design their own workflows. But then we throw it over the fence and hope for the best."

In the TSFS pod episode that just dropped today I was speaking with Brian Davies, VP at Safeguard Events and a well known voice in the industry. Coincidentally or not, we mostly talked about AI.

One of the topics we touched on was Brian’s observation on the difference between AI activation and AI enablement.

And it isn’t about semantics.

It’s a symptom of a general problem our industry is facing - what does using AI really mean? When can I say that I am using AI? And why does it even matter?

Let’s talk about it.

Unpacking the semantics

It’s September 2026 and it’s very hard to find an organization that isn’t using AI somehow, somewhere.

But there’s using AI, and then there’s using AI.

Fraud teams usually start where most teams start - in adopting AI for productivity gains. Transcribing calls, small automations over slack, and spending less time on writing reports and slides.

There’s of course nothing wrong with that, but are you really leveraging AI as a fraud fighter or as an employee?

In some organizations, what starts with a big announcement often ends up with the company spending huge dollars on tokens that end up not being used (other than for pet projects).

Even worse, some companies still measure token spend as a metric for efficiency/effectiveness, which often leads to vanity projects and unmaintainable AI slop.

The fact that budget has been spent, that every employee has an account, and that we told the board we’re now 100% AI-enabled doesn’t really mean anything.

It’s not until your team can take advantage of AI to positively impact their KPIs and goals in a measurable way, that you can say your team is AI-enabled.

And that’s the main split between teams that 4x their impact versus teams that only gain a 40% increase.

Overconfident inside, “imposter syndrome" outside

Just a couple of days before we recorded our conversation Brian published an interesting post on LinkedIn. One that made me chuckle.

He grouped together the responses he got from Safeguard attendees in regard to where their companies are on their AI journey. Here’s the result:

91%(!!!) of the responders rated their companies as AI-enabled, AI-first or AI-native.

Why did it make me chuckle?

Well, between Sardine and my own consulting practice, I get to see and speak to many companies across different geographies and verticals.

And let me tell you something, the real number is lower than 91%. Much lower.

Here’s the thing:

This overconfidence is actually a symptom of insecurity.

We live in an echo-chamber that is currently ruled by one story only: AI. No one is talking about how they’re not using AI. No one is talking about ML or rules, or processes.

So when you look outside, and you hear that 91% of the companies are at least AI-enabled, you think you’re behind.

That feeling of FOMO is what I encounter in each and every conversation I’m having.

And I fear that it also leads to overestimation of our own capabilities when we self-report.

Asking the right questions about AI

To me the question “are you using AI”, the one the board asks and ends up trickling down the company, is the wrong one.

And not only is it the wrong question, it’s also one that is likely causing you damage.

The reason is that when you’re busy thinking about whether you’re using AI, or to what extent, you’re not asking the right questions:

What do I gain by using AI today?
What’s my ROI on that gain?
What will increase my ROI?
How do I measure it all?
How do I maintain it?

If you don’t know the answers to these questions, it is highly unlikely that your “using AI” is actually leading anywhere or serving a purpose other than ticking a box.

But to ask these questions, you already need to be in a rather advanced stage. You didn’t only activate AI, you enabled your organization to create value with it.

How do you do that?

Brian’s four-point bar for enablement

Brian’s framework to self-analyze your AI activation level is pretty straightforward and is made of four points:

1) Policy. Employees need to know exactly what they can connect the tool to, and what they can't - email, CRM, payment processor, whatever's relevant.

Without a clear boundary, your most curious employees end up drawing that line themselves.

2) Training. Not a one-hour kickoff. People need to learn what a trustworthy output looks like versus a confident, wrong one, and that takes repetition.

3) Budget. Not just licenses, but time as well.

If using the tool properly stretches someone's day from eight hours to nine, you've outsourced your AI strategy to whoever works unpaid overtime.

Not the best strategy if you aim for a positive outcome.

4) Process. A recurring loop rather than a launch event - reviewed, adjusted, and reinforced on a cadence, managed the way you'd manage a person rather than a rollout you check off and forget.

Hit these four points, and you’re likely to start seeing real impact without defining a single project.

Because you’ve created a program.

The bottom line

It’s not about whether you use AI or not.

It’s about how you roll it out, how you enable your team to use it, how you measure it, and how you maintain and improve this motion over time.

To do that, you need to approach it programmatically, and not as a singular decision.

If you want to learn more about the journey of AI adoption, Brian had some really interesting takes on what you need to consider and how you can avoid some of the common pitfalls throughout it.

Do you have clear measurements on your AI program’s ROI? Hit reply and tell me - I'd love to compare notes.

In the meantime, that’s all for this week.

See you next Saturday.

P.S. If you feel like you're running out of time and need some expert advice with getting your fraud strategy on track, here's how I can help you:

Free Discovery Call - Unsure where to start or have a specific need? Schedule a 15-min call with me to assess if and how I can be of value.
Schedule a Discovery Call Now »

Consultation Call - Need expert advice on fraud? Meet with me for a 1-hour consultation call to gain the clarity you need. Guaranteed.
Book a Consultation Call Now »

Fraud Strategy Action Plan - Is your Fintech struggling with balancing fraud prevention and growth? Are you thinking about adding new fraud vendors or even offering your own fraud product? Sign up for this 2-week program to get your tailored, high-ROI fraud strategy action plan so that you know exactly what to do next.
Sign-up Now »

Enjoyed this and want to read more? Sign up to my newsletter to get fresh, practical insights weekly!

Chen Zamir