Influence of publicness dimensions on BD and AI activities
A table categorizes the influence of publicness dimensions on various B D, and A I activities.The table is divided into 5 columns and 7 rows. The first row depicts the column headers, from left to right, as follows: Column 1: Category; Column 2: Ownership; Column 3: Funding; Column 4: Goal Setting; and Column 5: Control Structure. The row-wise entries in the table are as follows: Row 2: Category: B D Collection; Ownership: Public ownership and supply chain integration enable data collection; Funding: Little funding is available for highly public organisations; Goal Setting: Data collection is limited without a clearly defined use case; Control Structure: Laws either facilitate or restrict data collection depending on the sector. Row 3: Category: B D Management; Ownership: Data stewardship remains with public institutions, but infrastructure is often outsourced; Funding: Limited funding is available unless infrastructure is critical to operations; Goal Setting: High-level goals often misalign with I T capabilities; fragmented responsibilities reduce strategic influence on data management; Control Structure: Strict laws govern privacy and data protection, setting requirements for B D management. Row 4: Category: B D Use; Ownership: Use is often restricted due to fragmented data ownership and legacy systems; Funding: Little dedicated funding, use typically occurs through specific projects or pilots; Goal Setting: Current processes don’t require significant B D use; public value and infrastructure efficiency drive use; Control Structure: Laws restrict data use beyond the originally intended scope. Row 5: Category: B D Sharing; Ownership: Lack of visibility in the supply chain and departmental ownership conflicts hinder data sharing; Funding: Ongoing funding is insufficient for interoperable platforms and open data initiatives; Goal Setting: Transparency and open government objectives encourage data sharing; Control Structure: Laws and regulations define and limit conditions for data sharing. Row 6: Category: A I Development; Ownership: Public A I projects face higher demands for transparency and ethical review, slowing progress; Funding: Budget constraints and risk aversion hinder long-term investment in A I systems; Goal Setting: Political agendas rarely prioritise backend I T modernisation, leading to fragmented initiatives; Control Structure: Legislation classifies A I applications into risk categories, imposing stricter requirements for high-risk A I. Row 7: Category: A I Application; Ownership: Public agencies often lack ownership of critical datasets required for A I model training; Funding: Efficiency demands support faster A I funding in less public entities; Goal Setting: Privacy risks and potentially significant impacts limit A I application in the public sector; Control Structure: Transparency requirements mandate explainable A I, especially in legal and social applications. Additionally, a legend at the top categorizes the entries as “Low Influence”, “Medium Influence”, and “High Influence”. The entries categorized as “Low Influence” are as follows: Row 2, columns 2 and 3; Row 3, columns 2, 4, and 5; Row 4, column 2; Row 5, column 3; and Row 6, column 2. The entries categorized as “Medium Influence” are as follows: Row 4, columns 3 and 5; Row 5, columns 2 and 4; Row 6, column 3; and Row 7, columns 2 and 3. The entries categorized as “High Influence” are as follows: Row 2, columns 4 and 5; Row 3, column 3; Row 4, column 4; Row 5, column 5; Row 6, columns 4 and 5; and Row 7, columns 4 and 5.
