Broadridge, a global Fintech leader, in collaboration with The ValueExchange, the International Securities Services Association (ISSA), and AI-specialist DeepSee, today released, AI in the Real World 2026: From AI Tools to Controlled Delegation in Post-Trade Operations. The report finds that 91% of respondents want providers to use AI to improve accuracy and reduce operational errors, followed by faster processing (85%), and proactive issue resolutions (80%). The findings put measurable service outcomes, rather than AI adoption alone, at the center of the industry’s next phase of investment. Drawing on more than 200 survey responses and interviews with market participants, the report finds that 61% expect moderate, significant or critical disruption if AI were removed, 78% say no more than one in ten AI use cases operates autonomously. The opportunity is to move from AI to controlled, accountable workflows, with trusted data, clear permissions, and human oversight. “The test of AI is not how many tools a firm deploys, but whether it delivers fewer errors, better client service and increased capacity,” said David Kirby, Head of U.S. Consulting Services, Broadridge. “Achieving those results takes more than a model. Firms need experts who understand post-trade workflows, trusted data and the controls that make AI dependable in production. Broadridge brings that domain expertise together with AI capabilities to help clients target the right opportunities, measure progress and scale with accountability.” The adoption gap in four numbers 49% report generative AI in production, compared with 11% for agentic AI.78% say no more than 10% of their AI use cases operate autonomously.70% describe asset servicing as fully or mostly manual; just 8% say it is fully automated.80% expect to use AI in external interactions by the end of 2027. Clients want fewer errors and better service Respondents say clients want practical outcomes from AI providers, not technology for its own sake. Improved accuracy and fewer operational errors lead the list (91%), followed by faster processing (85%), better data and reporting (83%), and proactive issue resolution (80%). For firms adopting AI, operational efficiency and cost reduction are the leading objectives, cited by 71% of respondents. The report argues that firms should also measure increased capacity, reduced rework and more consistent service delivery, not cost savings alone. Firms should establish a baseline for errors, rework, turnaround time and capacity, then evaluate AI against those measures. Broadridge combines post-trade expertise, transactional and operational data, and production-ready AI capabilities available through managed services or within clients’ own operations to help firms translate these priorities into practical workflow improvements. The next test: moving AI into core workflows Application development, client queries, and data quality lead current AI adoption. In asset servicing, 29% of respondents report current AI use and 47% plan to use it, while 70% still describe the process as fully or mostly manual. Settlement, fund accounting, and liquidity management also have more planned than live activity, indicating where firms expect the next wave of implementation. As firms move deeper into core workflows, systems must do more than generate an answer: they must use reliable evidence, act within defined permissions, document their actions, and escalate cases that require human judgment. “The strongest solutions combine AI with process certainty,” said Steve Shillingford, CEO, DeepSee. “In the middle and back office, a wrong answer is a settlement fail or a break someone has to chase. So, a system needs to show the evidence behind a recommendation, act only within its authority, and recognize when to stop and involve a person. When every step is verified against the firm’s own process, AI becomes dependable enough to take on more responsibility. That is the approach we have in production with Broadridge today.” Controls must keep pace with reliance The research identifies a gap between establishing AI policies and building the assurance needed to expand AI’s role. While 75% of respondents have a responsible-use policy and 65% report defined controls, only 46% report regular audits or explainability requirements, and 24% have processes for building stakeholder trust. The report describes the next stage as “controlled delegation”. AI performs bounded work and takes permitted actions, with clear accountability, monitoring, and escalation. Rather than pursuing autonomy everywhere, firms should expand AI’s authority selectively, based on the risk of each task and demonstrated performance.NoYesArtificial Intelligence17 Nov, 2026