Originally published on the IBM Banking Industry blog.

By Sonali Bhavsar, Associate Partner, Watson FSS Industry Platforms, IBM and Robert Cruz, Senior Director of Information Governance, Smarsh

FINRA and other regulators have informed firms of their expectation to adopt practices of “heightened supervision” to isolate high-risk brokers and activities, including those who have previously committed financial crimes. In the past two years, FINRA has released guidance to strengthen controls on high-risk brokers and further indicated it would be an ongoing priority. In addition, the Federal Reserve and a host of other regulators have released guidance on not only how to address the opportunities of artificial intelligence in financial crime as well as risk management, but what is expected once they look to apply innovative technologies.

Heightened supervision expands the aperture

FINRA’s 2018 Regulatory and Examination Priorities Letter expanded on a host of areas that institutions are expected to detect, prevent, manage and/or report, including fraud, high-risk firms and brokers, money laundering, cyber risk, liquidity risk, short sales, sales practices and suitability, initial coin offerings and cryptocurrencies, market manipulation and a handful more.

While the intention behind this guidance is much needed, and most financial institutions are already actively managing many of these areas for decades, often putting it to practice is a challenge across people, process and technology. Activities associated with financial crimes have become more difficult to identify and intercept given today’s volume, veracity and variety of communications data and the expanded focus also adds to the risk that firms won’t catch them all.

Moreover, these financial crimes are typically perpetrated outside of email, which has been the most scrutinized medium within the firm. But that is exactly why financial criminals choose alternate means of communication – because they know firms have limited insight across channels like voice, chat and messaging.

Expanding the focus beyond email surveillance

Looking beyond email is more than simply building a connector to an additional content source. Each new communication source is unique, some with well established methods of access and APIs, others not so much. Many new sources offer rich, dynamic, multi-modal functionality such as the ability to edit or delete content – or move a conversation across channels – which could result in losing information that is critical to conducting an investigation or responding to a regulator. Having context in an effective surveillance solution is an important factor for identifying true positives and negatives. When a typical financial services firm supports more than 40 different communications sources (and many firms support significantly more than 40), the challenges of capturing, preserving and supervising the breadth of today’s communications sources can easily grow exponentially.

Connecting the dots across channels and activities

The first reaction to this problem may be to create rules and protocols to limit the ability for employees to communicate, however, that may potentially increase the use of non-sanctioned communications channels. Instead, firms need the ability to not only bring all of these distinct communication channels together, but also establish a holistic approach that views communications in conjunction with employee activities, providing greater context and deeper understanding of what happened and why.

Historically, this has been the insurmountable challenge. However, artificial intelligence, machine learning techniques, and cognitive analytics provide the accuracy and efficiency to identify these incidents as they happen. As these techniques evolve with rich data and context, the identification process will only become more effective and efficient.

IBM Surveillance Insight for Financial Services takes a holistic approach to surveillance. It moves beyond the traditional rules-based alert detection, enabling firms to proactively monitor employees in real-time including emails, chat transcripts, voice recordings, trade and market data. In addition, greater agility and flexibility from more innovative technologies helps firms more quickly identity new predictive patterns for applications across conduct risk, market abuse and client suitability without a drawn-out model building process.

Getting a better perspective with holistic surveillance

Being able to tie together spoken and written words with corresponding actions is the only way to truly understand context and intent. With the mounting frequency and severity of insider and rogue incidents, spotting those red flags prior to the actual event is not just a nice to have; it’s exactly what FINRA and other regulators are asking for and it may make the difference between stopping a crisis before it happens and trying to figure out why you missed the warning signs.

Join the discussion

To learn more about FINRA’s “heightened supervision” guidance, watch our joint webinar with IBM, Financial Crimes & Focus on High-Risk Brokers, to understand the main obstacles to achieving compliance and examples of how to satisfy complex communications oversight.