Managing Change and Building AI Capabilities in Tax Teams
Expert-defined terms from the Professional Certificate in AI for Tax Technology Integration and Innovation course at UK School of Management. Free to read, free to share, paired with a globally recognised certification pathway.
Agile methodology #
A project management and product development approach that emphasizes flexibility, collaboration, and customer satisfaction. It involves iterative progress, continuous feedback, and rapid adaptation to changes. In the context of managing change and building AI capabilities in tax teams, agile methodology can help teams respond effectively to the dynamic nature of AI technology and tax regulations.
AI ethics #
The set of principles and guidelines that govern the design, development, deployment, and use of AI systems. AI ethics aims to ensure that AI is fair, transparent, accountable, and respects human rights and values. In tax teams, understanding and adhering to AI ethics is crucial for building trust, maintaining compliance, and avoiding unintended consequences.
AI model #
A mathematical or statistical representation of a real-world problem or phenomenon, designed to learn from data and make predictions or decisions. AI models can be used in tax teams to automate routine tasks, identify patterns and trends, and support decision-making. Examples of AI models include decision trees, neural networks, and support vector machines.
AI operations (AI Ops) #
The practice of managing and optimizing AI systems throughout their lifecycle, from development and deployment to monitoring and maintenance. AI Ops involves integrating AI systems into existing workflows, ensuring their performance, scalability, and reliability, and mitigating risks and challenges. In tax teams, AI Ops can help teams manage the complexity and diversity of AI technologies and applications.
AI strategy #
The plan and approach that organizations adopt to leverage AI for competitive advantage, innovation, and growth. An AI strategy should align with the organization's overall objectives, values, and risk appetite, and address issues such as data management, talent development, and ethical considerations. In tax teams, an AI strategy can help teams prioritize and focus their AI efforts, and ensure that they deliver value and impact.
Algorithmic auditing #
The practice of examining and evaluating the algorithms that drive AI systems, to ensure that they are fair, transparent, and accountable. Algorithmic auditing can help tax teams identify and address biases, errors, and inconsistencies in AI models, and ensure that they comply with legal and ethical standards.
Algorithmic transparency #
The degree to which the workings and outcomes of AI systems are understandable and explainable to humans. Algorithmic transparency is important for building trust, ensuring accountability, and mitigating risks in AI applications. In tax teams, algorithmic transparency can help teams understand how AI models make decisions, and communicate their rationale and limitations to stakeholders.
Change management #
The process of planning, implementing, and monitoring changes in organizations, to minimize disruption and maximize benefits. Change management involves engaging stakeholders, communicating vision and goals, building capabilities and skills, and measuring and evaluating outcomes. In the context of managing change and building AI capabilities in tax teams, change management can help teams navigate the complexities and uncertainties of AI integration and transformation.
Data governance #
The framework and processes that organizations use to manage and ensure the quality, security, and compliance of their data. Data governance involves establishing roles and responsibilities, defining policies and standards, and implementing controls and measures. In tax teams, data governance is essential for ensuring the accuracy, reliability, and privacy of the data used in AI models and applications.
Data literacy #
The ability to read, understand, and communicate data in context, and use it to inform decisions and actions. Data literacy is a critical skill for tax professionals in the age of AI, as it enables them to work with data-driven tools and methods, and make informed judgments based on data insights.
Data quality #
The degree to which data is accurate, complete, consistent, and relevant for its intended use. Data quality is a key factor in the success and reliability of AI models and applications, as it affects their performance, validity, and trustworthiness. In tax teams, ensuring data quality involves establishing data management practices, validating data sources, and monitoring data integrity.
Explainable AI (XAI) #
The branch of AI that focuses on making AI models and decisions understandable and interpretable to humans. XAI aims to bridge the gap between the technical complexity of AI systems and the human need for transparency and accountability. In tax teams, XAI can help teams build trust, ensure compliance, and make informed decisions based on AI insights.
Feature engineering #
The process of selecting, transforming, and scaling the input variables or features used in AI models. Feature engineering is a critical step in the AI model development process, as it can affect the model's accuracy, generalizability, and interpretability. In tax teams, feature engineering can involve selecting relevant data sources, creating meaningful features, and testing different combinations and representations.
Model drift #
The gradual degradation or change in the performance of AI models over time, due to factors such as data shifts, concept drifts, or model aging. Model drift can lead to inaccurate or biased predictions, and undermine the trust and reliability of AI systems. In tax teams, monitoring and mitigating model drift involves tracking model performance, updating data sources, and re-training models as needed.
Model interpretability #
The degree to which the internal workings and decision-making processes of AI models can be understood and explained by humans. Model interpretability is important for building trust, ensuring accountability, and mitigating risks in AI applications. In tax teams, model interpretability can help teams understand how AI models make decisions, and communicate their rationale and limitations to stakeholders.
Model validation #
The process of evaluating and assessing the performance and generalizability of AI models, to ensure that they meet specified criteria and standards. Model validation involves testing the model on new or unseen data, and comparing its outcomes to benchmarks or ground truth. In tax teams, model validation is essential for ensuring the accuracy, reliability, and fairness of AI models and applications.
Natural language processing (NLP) #
A subfield of AI that deals with the interaction between computers and human language, and the ability of machines to understand, generate, and translate natural language text or speech. NLP is a key technology for tax teams, as it enables them to automate and enhance various language-based tasks, such as document review, chatbot interaction, and sentiment analysis.
Reskilling and upskilling #
The process of developing and enhancing the skills and competencies of employees, to prepare them for new roles, tasks, or technologies. Reskilling and upskilling are important strategies for tax teams in the age of AI, as they help teams adapt to the changing landscape of tax technology and regulation, and maintain their relevance and value.
Shadow IT #
The use of unauthorized or unsupported IT systems, applications, or devices, by employees or teams, outside of the formal IT governance and control framework. Shadow IT can pose risks and challenges for tax teams, such as data security, compliance, and interoperability, and undermine the integrity and reliability of AI systems.
Tax technology stack #
The set of technologies, tools, and platforms that tax teams use to manage and automate their tax processes, workflows, and functions. A tax technology stack can include various components, such as data management systems, tax compliance software, AI models and applications, and analytics and visualization tools. In tax teams, building and maintaining a robust and integrated tax technology stack is essential for ensuring the efficiency, effectiveness, and innovation of tax functions.
Transfer learning #
The process of applying the knowledge and insights gained from training an AI model on one task or dataset, to another related or similar task or dataset. Transfer learning is a powerful technique for tax teams, as it enables them to leverage pre-trained models and data, and accelerate the development and deployment of AI models and applications, with less data and computational resources.
Unstructured data #
Data that does not have a predefined format or structure, and cannot be easily processed or analyzed using traditional methods or tools. Unstructured data can include various types of data, such as text, images, audio, and video, and can pose challenges and opportunities for tax teams, in terms of data management, analytics, and insights.
Workflow automation #
The use of technology and tools to automate and streamline business processes, workflows, and tasks, to improve efficiency, productivity, and quality. Workflow automation is a key application of AI in tax teams, as it enables them to reduce manual effort and errors, and focus on higher-value tasks and activities. Examples of workflow automation in tax teams include document management, data entry, and approval processes.