The news: The final report from the UK’s AI Public-Private Forum (AIPPF) points to the importance of data quality in artificial intelligence (AI) operations for financial institutions.
The AIPPF, established in 2020 by the Bank of England and the Financial Conduct Authority (FCA), was set up to facilitate dialogue between the private sector, public sector, and academia regarding AI.
Data comes first: The report describes data as foundational for AI, attributing most of AI’s recent growth to a surge in the availability of data to contribute to models.
- Data is more connected to AI's pros and cons than other aspects, and “many of the benefits and risks can be traced back to the data, rather than the AI systems or algorithms themselves.”
- Data is also among “the defining features of AI,” which can process massive quantities of data and find patterns from it.
What matters in data management? The report honed in on key areas for banks’ data operations for AI, such as:
Data quality: This covers accuracy, timeliness, transparency, and completeness.
- Banks’ quality challenges include needing to update their controls and processes for AI, and handling complex data sources.
- The report suggests that companies create their own quality assessment systems and standards, as well as integrating internal and external data sources.
Handing and risks: This includes determining the origins and legal status of data obtained from third-party providers that gather it from multiple sources and use website scraping.
- This involves challenges over liability and ownership of transferred data.
- Data assessments and audits can be beneficial, and questions about what data consumers will agree to share need to be considered.
- Implement data-sharing approaches that can help with privacy, like homomorphic encryption, which involves doing calculations on encrypted data without decryption, and differential privacy, where datasets’ characteristics are shared without naming individuals.
Data economics: Companies’ business models are shaped by the value placed on data, which can offer competitive advantages.
- Anticipated payoffs and data pricing should inform companies’ data usage.
- But cost-benefit analyses of data may be challenged when it’s third-party, and when dataset access is limited.
The big takeaway: Data is central to AI, and the forum’s report underscores the necessity of tackling data quality, privacy, and monetization. Banks can use these findings as a roadmap to improve their data operations, along with recommendations in our 2020 AI in Banking report, including:
Centralizing data to give business units using AI easy access.This can require updating technology infrastructure and data-handling processes.
Transparency over how customers’ data will be used. Users should be able to opt out of data collections. Intrusiveness about gathering and using consumers data could hurt their trust in and relationship with their banks.
Dedicate a sizable number of employees to data security. Investing in talent is key to reducing the risks of data breaches. Data security is critical to maintaining consumers’ trust.
Banks that get their data-handling operations right have significant consumer-product opportunities, including personalized insights, AI-powered biometrics and virtual assistants, and targeted offerings.