# How compliance leaders are guarding the truth in the AI era

> Source: <https://www.cityam.com/how-compliance-leaders-are-guarding-the-truth-in-the-ai-era/>
> Published: 2026-06-24 10:05:41+00:00

# How compliance leaders are guarding the truth in the AI era

The rapid acceleration of the digital landscape is presenting the financial sector with a dual challenge, where keeping pace has become about managing the sheer volume of data trailing in its wake.

Compliance and risk teams are currently navigating a huge shift. Teams are moving away from the traditional view of compliance as a back-office box-checking exercise, instead positioning themselves as key partners that enable firms to innovate safely. At the recent Smarsh Connect event in London, industry leaders and data privacy experts gathered to map out this transformation. The overriding consensus from the summit was the need to move forward confidently and to master an increasingly complex web of digital communications while maintaining corporate trust.

## The modern data challenge

While business communication was historically confined to straightforward, easily archived channels like official emails and landlines, modern employees now interact across dozens of different platforms, from instant messaging apps to collaborative tools. This fragmented ecosystem has generated a new wave of unstructured data, leaving compliance teams to sift through endless chats or audio files that cannot be easily captured by traditional databases. The biggest hurdle for risk officers lies in isolating these insights from the other digital noise. Faced with a relentless influx of automated notifications and email disclaimers, tracking down genuine compliance risks or market misconduct has become akin to searching for a needle in a haystack.

“The truth has got more rules these days,” said Smarsh chief executive Kim Crawford Goodman at the event. “And therefore the ability to actually see it, analyse it, know it, rely on the actual truth from real data is something that we very proudly do.” To address this bottleneck, forward-thinking institutions are reframing their approach to data archives. Rather than treating stored conversations purely as a regulatory burden, financial firms are beginning to leverage them as a rich source of context. Goodman explained that international firms are actively seeking ways to extract deeper meaning from their archives: “What I need from you is for you to pull intelligence and insight that will put context around all of that structured data. And in fact, what we’re seeing is… it’s the context that adds the intelligence.” Recognising that understanding the underlying intent of a conversation gives risk teams the necessary edge to mitigate threats before they escalate into bigger, more dangerous failures.

## Managing evolving threats

As digital communication tools have outpaced legacy systems, many organisations have accumulated significant “data debt” by falling behind in their management strategies. Relying on blanket bans on specific applications or implementing overly rigid corporate policies has proven ineffective for securing a firm’s scope beyond its corporate firewall. Sonia Cheng, a Senior Managing Director at FTI Consulting, argued for a shift in how companies approach these blind spots. “Companies can no longer rely on prohibition policy as a means of safeguarding against activities and messaging taking place outside of corporate firewalls,” Cheng said. “They need to think holistically about risk. It’s really about thinking pragmatically about where the materiality of the risk is, and then tackling that specifically. In other words, don’t try to boil the ocean.” Deploying advanced analytics or AI on top of compromised, unorganised data repositories is an operational dead end, because if the underlying data foundation is weak, the technology will inevitably fail to deliver accurate results.

Cheng said that the shifting landscape makes legacy weaknesses particularly dangerous: “It’s less about mythos and more about the fact that it’s a moving target. As computing power increases, you have new tech that is better able to target vulnerabilities in your code and your data.” There is also a concerted push among risk leaders to prevent a governance gap from opening up around newer tools like generative AI. Just a decade ago, the proliferation of mobile messaging apps left the financial sector scrambling to adapt to shifting regulations, and the objective now is to remain ahead of the curve. Goodman said: “We do not want this industry, for AI, to become the next off-channel communications where people are using it, using it, using it; value, value, value; customers are impacted, and you’re behind the eight ball. Then the regulator has to show up… No, we’re way in front. We’re way in front of that.”

## Collaboration has become key

Tackling these modern data challenges requires an overhaul of corporate software systems, from traditional, legacy setups that merely recorded historical events to intelligent platforms that filter data at the point of ingestion. These advanced systems automatically filter out irrelevant spam and highlight useful communication patterns before data enters the archive, drastically reducing data noise. However, the tech alone cannot solve a cultural problem, and successfully upgrading an enterprise requires a balanced corporate mindset. Organisations frequently fall into the trap of either trusting a new tool blindly or dismissing it entirely because an early iteration made an error. Cheng said: “It’s the mindset discussed at the conference that I think is so important: that growth mindset… being willing to embrace failure, fail fast, trying those things, but trying safely. Getting that balance of commerciality and innovation with the risk lens.” Ultimately, achieving responsible innovation depends on having a continuous conversation between commercial teams, legal experts, compliance officers, and cyber teams. This approach ensures that while a company aggressively develops new digital experiences for its clients, it does so without compromising data privacy, security, or regulatory requirements.

## Pressure to progress

The commercial reality of upgrading systems has shifted dramatically in recent years, as processing costs for running smart AI tools have dropped significantly, making the tech accessible to businesses of all sizes. Consequently, a passive wait-and-see approach to technological adoption is no longer viable for financial institutions. Goodman argued that the industry’s baseline has shifted quickly, contrasting her advice from just 12 months ago with current market pressures. “A year ago… what I was saying is, as the leader in compliance communications, we will be there for you whatever you need… If you want to just get by your next exam and have maximum amount of explainability, and you’re not really on the AI train, we’re here for you as well. That’s what I was saying a year ago.” But now that passive stance is obsolete, meaning risk leaders need to protect their data and support their teams in maintaining market relevance. “I’m saying something completely different this year,” Goodman added. “I’m saying that everybody in this room, everybody in our Smarsh community, needs to be meaningfully on this train… Nobody here is going to be driving the horse and buggy while the rest of the world is off driving Ferraris and F-16s. It is now time to pick up and go.”
