'Fake data' will make banks vulnerable - Accenture

'Fake data' will make banks vulnerable - Accenture

Many banks are not doing enough to verify the validity and accuracy of their data, leaving them vulnerable to false insights that could lead to bad decisions, according to Accenture.

As financial services firms grapple with ever-growing mountains of data, often from external, unstructured sources, a poll of nearly 800 bankers by Accenture suggests that half aren’t doing enough to validate and ensure data quality.

More than nine in 10 of the bankers surveyed say they are confident in the integrity of the sources of their data. However, 11% trust their data is reliable, but don’t validate it; 16% try to validate their data, but aren’t sure of the quality; and 24% validate the data, but recognise they should do a lot more to ensure the quality.

In addition, while five in six bankers say they increasingly use data to drive critical and automated decision-making, more than three-quarters believe that these automated systems create new risks, such as fake data, external data manipulation and inherent bias.

Alan McIntyre, head, banking practice, Accenture, says: "Given that four in five bankers that we surveyed said they are basing their most critical systems and strategies on data, it’s critical that the data can be verified and validated."

The survey also shows that AI and the blockchain continue to be top topics of interest for the industry. On average, respondents expect that operational blockchain systems will be live in their banks in 2.6 years, and that the technology might one day provide cost-effective replacements for legacy core banking systems.

Nearly four in five bankers believe that AI will work alongside humans as collaborators and trusted advisors within the next two years, although many are worried about regulatory and ethical standards.

Peter Sidebottom, an MD of strategy in Accenture’s FS practice, says: "AI decisioning processes can’t be a black box; banks need to adhere to and provide the same transparency as they do with any other employee to ensure regulatory compliance, and to earn customer trust."

Read the full report

Download the document now 1.6 mb (Chrome HTML Document)

Comments: (2)

Bo Harald
Bo Harald - Transmeri, Demos, Real Time Economy Program,MyData - Helsinki Region 22 April, 2018, 10:44Be the first to give this comment the thumbs up 0 likes

Banks are in very central positions when data is needed for a long range of contexts for human customers (Ms Same Guy in private and enterprise roles..). Banks have the best starting point - trust - in the data economy. But they need to come together to create the supporting ecosystems needed. With KYC in the legislation, PSD2, GDPR etc it is very difficult to understand why so many are sitting on their hands - starting from offering e-ID services with the bank e-Id...


Sheena Clark
Sheena Clark - Financial Machineries - London 23 April, 2018, 20:07Be the first to give this comment the thumbs up 0 likes


Banks need to look at their own data creation and publishing processes first as they are often the original price source, especially for OTC securities.

What data you contribute to which vendor and how is crucial to understanding regulatory and reputational risk as well as taking the highground in being able to ask the aggregators/vendors to produce the best (verifibale) rates possible that are up to your standards.

If you think the main data providers are working on your behalf, just ask them to:

Provide an instrument inventory of what you (as a global organisation) are contributing to them?

Ask to see which of your competitors has access to your data per type of permission?

Will these answers comply with MifID2?.....to be continued....