“We are basically aiming to do your tax return for you,” Bill Gibson, CIO of the Australian Tax Office told the Technology in Government Summit.
This is already happening to some extent, with the online lodgement system pre-populating taxpayers’ returns with information about salaries and wages, interest, and other data collected by the ATO from various sources.
But just as cloud accounting software can automatically process transactions appearing in bank feeds, the ATO is looking to do something similar for work-related expenses. Mr Gibson said the Commonwealth Bank already allows customers to tag items as being tax deductible, and the ATO needs to be open to accessing such data sources.
“It’s very exciting,” he said, noting that the ATO will need outside help to implement such ideas.
Asked whether data governance issues can lead to projects fading away, Mr Gibson said it was currently hard enough to get data from within the ATO, and across the whole of government it was about two orders of magnitude more complex. What that means in terms of accessing data from non-government sources (as opposed to data that must be supplied to the ATO) was left to the imagination.
Denise McDonagh, CTO at the UK Home Office (the government department responsible for crime, justice and policing; border security; national security; and rights) told the summit that there is “a constant tension” between maintaining citizens’ trust and usefully sharing data between agencies that had to be balanced.
The UK Government has very clear information assurance guidelines, and permission to share certain data is granted by some legislation. However, she believes people sometimes misunderstand the legislation or guidance, and do not really understand what they may or may not share.
A problem with data sharing is veracity, Mr Gibson explained: does the data really represent what it purports to represent? This is important because recipients of data will make good-faith decisions based upon it, he said.
Mr Gibson also said the ATO was not only thinking about how it can use relevant data from other sources, but also whether it should provide depersonalised data for outside use. But as Amr Awadallah (CTO of big data vendor Cloudera) told CSO, anonymised data can sometimes be combined with other data sets to reveal personal information.
A recent example was that supposedly anonymised data about 173 million taxi trips in New York has been processed to reveal drivers’ identities and potentially allow estimates of their gross incomes and approximate addresses.
To hear more from Stephen Withers watch this space... (To connect with Stephen Withers LinkedIn)
This article is brought to you by Enex TestLab, content directors for CSO Australia.