In the old days of hand made goods, the brand was enough to show that the craftsman had tested the good and certified its quality. As the scale of operations grew, the producers decided to retain the quality marks, but resorted to sampling(around which a whole cottage industry grew of sampling statistics, testings, Six Sigma etc). When the scope of financial audit enlarged from balance sheet audit to include transactional testing, test of controls etc; auditors just lifted the ready made library of tools/techniques from the sampling industry, under the mapping that transactions/records were like goods-for which 100% testing was not economically feasible. While that hypothesis is shrinking in physical goods with the advent of non destructive testing(like acoustic testing, use of optics etc), digitization has made it possible to apply CAATs and digital auditing to do 100% testing of information.
Technological feasibility is one reason, but the real reason is the increasing competition+globalization, which has really reduced profit margins. Companies must now innovate(to maintain the high margins/market share), or else die. The lower profit margins lead to lower audit materiality levels, and therefore need for more audit testing. After all, commodity players operating at 1% margins, would have low audit materiality levels. But even high margin financial sector players have other issues like lax controls, fraud risk, compliance testing etc. Money laundering legislations demand data mining to detect connected transactions. The stray transaction which slipped through, may result in headlines about XYZ bank being a terrorist conduit. To protect their reputational capital, those subject to money laundering legislation must have digital records. And when the underlying digital records are in place, it is a natural leap to use them for statutory financial audit purpose.
Another reason is the increasing frauds/collapse of listed companies, where the public/regulators blame the auditors for being asleep at the wheel. While this perception is due to expectation gap, the audit firms would naturally like to improve the audit quality to avoid such issues. And one of the cheapest(given their economies of scale across clients) ways to do so is to use computers effectively for 100% audit testing.
As I blogged earlier, the data analysis trend today is to use the automated bulldozer instead of the manual shovel. Hence, the need for smart work