||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| Regulation
Finding the data key
Establishing an industry regulator is the key to unlocking synthetic data’s potential
Stephen Brown Managing principle, Capco
The FCA’s consultation on ‘Synthetic data to support fi nancial services innovation’ points to the challenge of truly replicating real data sets and believes establishing a central regulator is key to unlocking synthetic data’s potential.
Valuable bridge Synthetic data can provide a valuable bridge for fi nancial service providers to gain insights into customer trends and preferences without breaking privacy and protection rules. However, achieving synthetic data accuracy
is critical to establishing the necessary trust within the industry to propel its adoption.
Immature As the use of synthetic data is largely immature, a lack of trust in its accuracy remains the biggest barrier to wider adoption. Where synthetic data leaves scope to be
interpreted as unrepresentative or patently false – for instance when used in conjunction with real data – this will likely generate invalid insights at best, and litigable falsehoods at worst.
Ultimately this will push fi rms to instead
favour alternative techniques, such as real world data anonymisation or pseudonymisation.
Accuracy issue To solve this accuracy issue, fi rms should ensure synthetic data is generated from real world, customer data provided by incumbent organisations, via processes that meet a regulator-defi ned set of standards. However, given that those real world
datasets provide the ‘data owner’ organisation with a competitive advantage over their
June 2022
Where synthetic data leaves scope to be interpreted as unrepresentative or patently false – for instance when used in conjunction with real data – this will likely generate invalid insights at best, and litigable falsehoods at worst
www.CCRMagazine.com In Focus
competition, many will likely be unwilling to volunteer their data for now.
More feasible A preferable and more feasible approach could be an FCA-approved standard that
would allow an organisation to take its own data and create its own synthetic datasets for use in its own projects. This achieves the goal of driving greater
adoption of the use of synthetic data at scale within an organisation. For the business, there is trust in that the
synthetic data is representative; and from a compliance perspective, there is mitigation of risk in that the synthetic data meets a certain set of regulator-defi ned standards.
Cross-collaboration Cross-collaboration with other regulators will also be fundamental to establishing standards for generating synthetic data from an organisation’s own data. Without it, widespread adoption would
likely fail as the investment to create locale- specifi c synthetic datasets would represent a high bar of investment. CCR
29
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52
