Only 5.2% of Tech Leaders Confident in Explaining AI Decisions, FICO Research Reveals
By Lauren Towner · 12 October 2026
In brief: FICO's 2026 report reveals only 5.2% of tech leaders can fully explain AI decisions, despite 85% seeing positive ROI, highlighting a critical need for Responsible AI standards.
| Survey Size | 1,004 senior leaders |
|---|---|
| Confidence Level | 5.2% very confident |
| AI ROI Success | 85.1% met or exceeded expectations |
| Agentic AI Adoption | 2.4% broadly deployed |

FICO’s 2026 State of Responsible AI study reveals a significant gap between investment and implementation, with 71.5% of customer-impacting decisions still relying on legacy processes despite high ROI expectations. For fintech leaders, this highlights a critical bottleneck: a lack of explainability and standardized frameworks is preventing the full-scale deployment of automated decisioning across the enterprise.
What was announced
The fifth annual report, developed in collaboration with Corinium Global Intelligence, surveyed 1,004 senior technology, risk, and data leaders to track the adoption of responsible AI practices. Despite billions of dollars in investment across the sector, the study found that only 5.2% of respondents feel very confident explaining their AI-driven decisions to regulators or customers today. This lack of transparency is a primary barrier to adoption, as 37.3% of leaders admit they are neutral or not confident at all in their ability to justify automated outcomes.
The research indicates that while 85.1% of organizations report that AI has met or exceeded their ROI expectations, the majority of enterprise decisions still depend on legacy processes: humans (33.3%), rules-based systems (21.4%), and standard analytics (16.8%). To bridge this divide, 71.9% of respondents identified AI model development standards as the practice that would make the most significant difference in their operations. Furthermore, 67.9% of leaders believe interpretable model architectures are essential for the technology to reach its potential.
Infrastructure remains a significant hurdle to scaling these technologies. Only 8.2% of organizations have a single shared AI deployment platform fully in place, while 71.2% have one only partially in place. The emergence of agentic AI—autonomous systems capable of acting on behalf of users—is currently limited, with just 2.4% of tech leaders having broadly deployed it for customer-facing use cases. Most organizations (71.5%) remain in the pilot or exploration phases, citing security concerns (40.1%) and data integration challenges (38.3%) as the main obstacles to achieving ROI with agentic systems.
"Most organizations can tell you where and what their AI decided. Far fewer can tell you why. This points to a lack of responsible AI standards and techniques, which directly impacts organizations' willingness to use AI in customer-impacting decisions. As regulators and customers scrutinize how AI is used in decision-making, auditability and explainability are nonnegotiable."
Scott Zoldi, chief analytics officer at FICO.
The companies involved
FICO (NYSE: FICO) is a global leader in analytics software, best known for its credit scoring models that have become a standard in the financial services industry. The company provides a range of tools for decision management, fraud detection, and credit risk assessment. Scott Zoldi, the company's chief analytics officer, also serves as a member of the CNBC Technology Executive Council. Rachael Hadaway serves as the VP of AI Product at the firm. FICO has increasingly focused on "Responsible AI" as a core part of its platform strategy, aiming to provide the transparency required by highly regulated sectors.
Corinium Global Intelligence is a specialist market intelligence firm that focuses on the C-suite, particularly within the data, analytics, and digital transformation sectors. The firm produces high-level research and executive events designed to connect technology leaders with industry insights. Lariza Carrera serves as the Production Director at Corinium. Together, the two organizations have tracked the evolution of AI maturity over the last five years, documenting the shift from experimental technology to a regulated enterprise requirement.
What FF News has reported before
FF News has closely followed FICO’s efforts to integrate advanced analytics into the financial ecosystem. In July 2026, the publication detailed how FICO Launches New DataOps Capabilities to Accelerate Enterprise AI Adoption and Decision Intelligence, a move aimed at streamlining the very infrastructure gaps highlighted in the latest report. Prior to this, coverage focused on the company’s market reach, such as when FICO Adds Powerful Credit Optimization Tools to AWS Marketplace in late 2025. The firm’s research has also been instrumental in tracking criminal trends, as seen in FICO Research: AI Arms Race Drives 25% Surge in EMEA Fraud Attempts and the FICO European Fraud Map: UK Resilience Shines as Continental Card Fraud Hits €1.69 Billion.
What this means
The findings suggest that the "black box" nature of AI is no longer just a technical hurdle but a commercial liability. As the industry moves toward agentic AI, the pressure on financial institutions to provide audit trails will only intensify. The fact that over 70% of customer decisions still bypass AI indicates that trust remains the primary currency in fintech, not just processing power. This announcement puts vendors of proprietary, opaque models under pressure; the market is clearly pivoting toward interpretable architectures. The open question for the sector is whether standardized development frameworks can emerge quickly enough to satisfy regulators before the current wave of AI investment loses its momentum.
Frequently asked questions
Why are tech leaders struggling to explain AI decisions?
According to FICO, only 5.2% of leaders feel confident explaining AI decisions due to a lack of responsible AI standards and techniques. While organizations can track what was decided, many lack the auditability and explainability tools required to understand the 'why' behind the data, which is increasingly demanded by global regulators.
What is the current ROI for enterprise AI deployments?
FICO's 2026 study found that 85.1% of senior leaders say AI has met or exceeded their ROI expectations. However, despite this financial success, 71.5% of customer-impacting decisions are still made using legacy processes like rules-based systems or human instinct because of trust and transparency issues.
What are the main barriers to adopting agentic AI?
Agentic AI adoption remains low, with only 2.4% of organizations deploying it broadly. The primary barriers identified in the FICO report are security concerns (40.1%) and data integration or quality challenges (38.3%). Most leaders (71.5%) are currently keeping agentic AI in the exploration or pilot phases.
About the companies
Corinium Global Intellegence
Website · 1 story on FF News since 2026
Companies in this story: Corinium Global Intellegence, FICO
People in this story: Rachael Hadaway, Scott Zoldi, Lariza Carrera