A clear split has emerged between the sell-side and the buy-side in their approaches to enterprise data strategy, with banks adopting a more defensive posture and asset and wealth managers more aggressive in their pursuit of modern capabilities.
The divergence was highlighted in a recent A-Team Research and Element 22 webinar, entitled “The Data Office at a Crossroads — AI Governance, Organisational Design, and the Evolving Mandate of the CDO”.
Large tier-one banks, heavily influenced by a decade of strict regulatory mandates such as BCBS 239, continue to view data primarily through a defensive lens, panelists on the webinar agreed.
Those organisations’ operational focus remains on governance, structural control and quality assurance, often resulting in larger, more centralised teams with a propensity to build proprietary technology.
Conversely, asset and wealth managers operate with leaner models and exhibit a stronger orientation towards advanced analytics, with one panelist noting that the buy-side is leveraging a “last mover advantage”, unburdened by legacy infrastructure and therefore free to adopt modern capabilities more rapidly.
Ultimately, these approaches represent different points on an evolutionary continuum. As artificial intelligence capabilities mature across the industry, the rigid centralisation of defensive governance is expected to federate, pushing analytical capabilities directly into business lines.
The Fragmentation of Data Leadership
The webinar was the latest in A-Team Group’s series of events providing community-led intelligence for the data office. The series is presented in conjunction with data management specialist Element 22.
The panel comprised Mark Davies, partner at Element22; Duncan Cooper, founder of Duruedma Limited and formerly the chief data officer of Northern Trust; and Brian Greenberg, senior director – business engagement lead for enterprise data management at Bank of New York. The webinar was moderated by A-Team Group president and chief content officer Andrew Delaney.
The rapid integration of AI has triggered significant friction within corporate hierarchies, most notably manifesting in the fragmentation of C-suite titles, the panel noted.
More than half of the surveyed organisations now maintain both a Chief Data Officer (CDO) and a Chief AI Officer (CAIO) as independent functions. This bifurcation often stems from differing mandates: the CDO is traditionally tasked with maintaining stability, business-as-usual operations and mitigating risk, whereas the CAIO acts as a change agent deployed to actively disrupt the status quo.
However, this separation is probably a temporary symptom of technological immaturity, the panel stated. AI is inherently dependent on data to identify patterns and generate actionable insights. Consequently, as the enterprise understanding of AI evolves from focusing on underlying compute infrastructure to its practical business applications, these leadership roles are expected to re-converge.
The data office must actively drive AI strategy rather than serve merely as a passive participant or a compliance constraint.
Value Perception and Zero-Based Scrutiny
Securing budget and demonstrating a return on investment remain the most persistent challenges for data leadership, the panel agreed.
The initial honeymoon phase of unconstrained AI investment is closing, replaced by intense executive scrutiny. As one panellist highlighted, a growing number of financial firms are instituting zero-based budgeting for their data functions, demanding that leaders justify every line item and proposed outcome from scratch.
This requires a strategic shift from low-effort efficiency plays, such as automating meeting summaries, to high-value strategic growth initiatives. These advanced applications include optimising balance sheets, performing complex asset allocation, and identifying geopolitical risks buried in unstructured documentation. The data office must articulate an outcomes-based narrative, or risk having its funding permanently reallocated to more agile business units.
Addressing the Skills Deficit and the Rise of Digital Employees
Despite the availability of advanced tooling, the most significant underinvested risk in the financial technology sector is the AI skills gap. A vast majority of firms offer no AI-related training to their existing staff, the panel noted.
Addressing this requires a cultural shift where AI is integrated into the daily workflows of all employees, rather than being siloed within a central engineering hub. The industry is already seeing the successful deployment of “digital employees” – AI agents assigned to specific technical tasks, such as resolving code errors – operating under the direct supervision of human managers. While this human-in-the-loop model satisfies compliance and risk requirements, achieving scalable, autonomous value will eventually require profound structural changes across human resources, audit and legal frameworks.
The modern data office cannot afford to remain static, the panel agreed. The deep integration of AI demands a transition from defensive record-keeping to offensive value generation.
Institutions that cling to legacy architectures and protective data hoarding will inevitably lose their competitive advantage to more agile peers. By investing in comprehensive staff training, unifying data and AI leadership, and decisively reallocating budgets towards high-value analytical outputs, data leaders can navigate this crossroads effectively.
The mandate is clear: the data office must evolve from an operational participant into a strategic driver of AI, or it risks obsolescence in a rapidly accelerating market.
Subscribe to our newsletter
