Impact of

Artificial Intelligence (AI)

on Audit – An Introduction

By
K. Atique –e- Rabbani, B Tech (Hons), UK, FCA
 Impact of AI on Audit – an Introduction
Index
1. A word about me
2. Caution about AI
3. What is AI?
4. AI v NI – Natural Intelligence
5. History of AI
6. AI Evolution
7. AI in your phone – Siri
8. CAPTCHA
9. Stored Program Concept to Machine Learning
10. AI is a coordination of Machine Learning, Deep
Learning, Big Data, Cognitive Computing
11. IBM Watson – Cognitive Computing Platform
12. Application of AI
13. AI in Accounting
14. AI in Auditing (2 slides)
15. Blockchain and Bitcoin
16. Blockchain in Auditing
17. Closing thoughts
 A word about me
• I am a FCA.
• I also have a B Tech (Hons) degree from
University of Bradford, UK.
• I studied Statistics, Computer Science, Pure
Maths, Applied Maths., Numerical Ananlysis,
Operations Research.
• Then I trained as a Trainee Chartered
Accountant with Arthur Young McClelland
Moores & Co ( now Ernst Young) in Luton, UK.
• I am a Chartered Accountant who has been a
Programmer, Software Developer, Integrated
Technology Solution Provider, ICT Project
Manager and an ICT Consultant for over 33
years.
• I have not been a practicing Accountant as
such.
 Cautions about AI
Humans need resurrection

• Technology is good if we are good. • Alphabet/ Google uses neural

• If we don’t shape up AI could spell networks, for tasks such as
doom. • enabling self-driving cars to recognize
• Stephen Hawkins remarked that AI or objects,
thinking machines could, when fully • translating languages,
developed, spell the end of human race. • adding captions to YouTube videos,
• Elon Musk of TESLA said AI could be • diagnosing eye disease, and
more fatal than nuclear weapons – an • even creating better neural networks.
immortal dictator.
• Sergei Brin, Google Cofounder, with • Google got its name from googol
greatest investment in AI warns about which means 1 followed 100 zeroes.
AI.
• Google uses custom AI chips that are • Google’s quantum chips could one
million times more powerful than day offer speeds of ‘googol’ over
Pentium II chips. present computing power
 What is AI?
• Artificial intelligence (AI) as the
name implies is artificial.
• Intelligence that we and animals
have is natural intelligence.
• To understand natural intelligence
we have to fully understand how
our brains function.
• There are 100 billion neurons in
our brain.
• How brain works is still a big
mystery.
 AI v NI – Natural Intelligence
• Artificial Intelligence is apparent
“intelligence” demonstrated by
computers/machines.
• Apparent because true intelligence
is what us, mysterious living being
only have.
• From early day attempts were
made to understand whether
thoughts and reasoning were all
systematic like algebra.
• In early days artificial men and
thinking machines were depicted in
fiction.
 History of AI
• Artificial Intelligence is also not new.
• Alan Turing, one of the fathers of AI came
out with a concept called Turing’s test, in
1950.
• If a human cannot tell whether it is
talking to a human or machine over
teletype then that machine would pass
Turing’s test.
• The term Artificial Intelligence was coined
63 years ago in 1955 when both Bill Gates
of Microsoft and Steve Jobs of Apple
were born.
• John McCarthy who taught Mathematics
at Dartmouth College, Hanover, New
Hampshire, USA coined the term.
• Later McCarthy formed AI Labs in MIT and
Stanford and created LISP Computer
Language for AI programming.
 AI evolution
• Abacus for doing arithmetic
dates back to prehistoric times –
5th century BC.
• Slide rule invented in 1622 used
by engineers
• When IBM PC was introduced on
Aug 12 1981 they had no idea
what they had done.
• AI is an existential threat to
humanity -Elon Musk of TESLA.
 AI in your phone - Siri
• Robots or machines have not quite
taken over yet.
• At least for now.
• But times are not far off when we
will find them everywhere doing
most of what we humans were
supposed to be doing.
• From personal digital assistants like
Siri in IPhones, Alexa from Amazon,
Google Assistant to report writing
AI robots we are in for a new
experience here.
CAPTCHA
 Stored Program Concept to
Machine Learning

• Computers/ Machines traditionally store • The term ‘Machine Learning’ was coined
and follow instructions. in 1959 by Arthur Lee Samuel.
• John von Neuman in 1940s introduced • Machines progressively improve
this stored program concept. performance through feedback.
• Programmers program the computer to • Feedback is given through patterns and
do all the wonderful things, like trends.
Accounting and ERP that the computer • This discerning of patterns or trends is
does for us. done by data mining on Big data.
• Machine Learning, however, is a slightly • Big data as distinguished from data is
different concept from stored program extremely huge in volume and contains
concept. structured and unstructured data.
• Computers not only follow set of • Data mining techniques extract patterns,
instructions to do things repetitively but trends from Big data, say for financial
they can also ‘learn’ to do things better. transactions.
AI is a coordination of Machine Learning, Deep Learning,
Big Data, Cognitive Computing

• Artificial Intelligence is a
coordinated system which
involves Machine Learning, Deep
Learning, Big Data, Cognitive
Computing and much more.
• Artifical Intelligence (AI) uses
Intelligent Agents (IA) who can
sense and act accordingly.
• For example a thermostat is an
example of an Intelligent Agent
(IA).
 IBM Watson – Cognitive Computing platform

• IBM Watson is a cognitive computer platform • KPMG has partnered with IBM Watson to
that is being used across the world to boost integrate AI technology with taxes and
revenue, efficiency and even save lives. advisory services.
• With astronomical increase in computational • KPMG and IBM Watson has developed data
power and coming of age of human like analytics tool to address Lease Accounting
question and answering system like IBM and find trends, patterns, anomalies and
Watson AI is finding its use in our every day relationship in lease contract data.
life. • The above tool is for new IFRS 16 Lease
• Cognitive computing simulates human compliance (effective from 1 Jan 2019)
thought process in a computerized model. where Lessee treats pretty much all leases
• Cognitive computing consists of self learning in the same way
systems that mimics the way human brain • The Big Four PwC, EY, KPMG, Deloitte are
works. ramping up investment in AI and related
• This it does through data mining, pattern technologies to defend market share and
recognition and natural language processing. to address regulators’ tough stance at the
profession.
 Applications of AI

• AI is ready for prime time now. • Artificial Intelligence beats human

• It has been stuck in research labs for pathologists at predicting patient survival
decades. times for some cancerous tumors.
• AI can transform and disrupt all the • GO is an abstract strategy board game
sectors of the economy from tech to significantly more complex than Chess.
finance, communications, energy, • It was invented by Chinese 2500 years
healthcare, mobility or manufacturing ago.
that generate lots of data (big data). • Artificial Intelligence programme called
• Image Processing – Google’s employees AlphaGo beat decisively one of the
protested company’s contract with strongest GO players, Lee Sedol in 2016.
Pentagon to apply machine learning to • The rise of artificial intelligence has
videos from drones recently led to bots writing real news
• Artificial Intelligence calculates insurance stories about sports, finance and politics.
payouts in Japanese Insurance firm
better than humans.
 AI in Accounting
• Big data brings a new paradigm for accountants. Newer and
newer relevant data becomes available.
• The video, audio, and textual information made available via
Big Data can improve managerial accounting, financial
accounting, and financial reporting practices.
• In managerial accounting, Big Data will vastly improve effective
management control systems and budgeting processes.
• In financial accounting, Big Data will bring forth relevant
accounting information, thereby enhancing transparency and
stakeholder decision making.
• In reporting, Big Data can assist with the creation and
refinement of accounting standards, helping to ensure that the
accounting profession will continue to provide useful
information as the dynamic, real-time, global economy evolves
• Currently AI and Machine Learning are automating book
keeping, reconciliations and much more. Major accounting
companies like Intuit, Sage, XERO, OneUP, QuickBooks Online
are offering such capabilities from their cloud-accounting
platform. AI will be disruptive in the accounting arena. It will
take away time consuming and hence lucrative work from
human Accountants.
 AI in Auditing

• Auditing to be a less of a burden as • Auditors can harness the power of data mining
accountants embrace AI and AI to help avoid further high-profile
• Auditors use drones to use AI and image failures.
recognition to analyze inventory information • Automating some repetitive manual tasks will
• PwC and EY are using AI tools to automate part improve efficiency, by freeing up employees to
of the audit process concentrate on areas where human judgment
is needed.
• Firms are now exploring more sophisticated • Mindbridge AI is an analytics company using AI
artificial tools that “learn” over time the more to uncover material irregularities in data and
information they gather, gradually gaining the has built in AI Auditor, a world’s first.
ability to recognise complex patterns.
• Innovations in the pipeline include systems that • The MindBridge Ai Auditor analyzes all of the
can detect anomalies across all of a company’s transactions using a hybrid of artificial
financial transactions, such as unexplained intelligence, Machine Learning, statistical
entries that do not reflect what the business models and rule-based tests offering higher
normally does. assurance. This has been proven to be at least
10x more effective than current approaches.
 AI in Auditing (Contd)

• Audit as we now have using samples will • Cognitive Technologies, AI will enhance
not exist audit quality, automate processes and
• Web robots also known as bots transform how an audit is performed.
can perform simple and repetitive tasks at • Natural language Processing, data
breakneck speed analytics, machine learning are
• Bots going around the system will do automating and streamlining labour
continuous audit intensive processes.
• Bots can learn and pick out the abnormal • Auditors will focus on higher value added
patterns areas, provide insights and deliver value
beyond compliance.
• Good Bots gather good information
• Organizations are already using Ethics and
• Bad bots produced as those used in US Compliance Software such as
elections to launch malicious and harsh CONVERCENT, PREDICT360, iSight Case
attacks. Manangement Software, Navex Global
GRC (Governance, Risk, Compliance)
software.
 Blockchain and Bitcoin

• Blocks refer to record blocks and all blocks are linked using cryptography.
Hence block chain.
• Blockchain is a continuously growing list of records, called blocks.
• Each block contains a cryptographic hash of the previous block, a time
stamp and transaction data.
• There is no central authority controlling blockchain.
• It can record transaction between two parties in a verifiable and
permanent way
• For use of Blockchain as decentralized, distributed public digital ledger it
is typically managed by a peer to peer network
• No record can be altered retroactively over a network of computers
without the collusion of network members.
• A global network of computers uses blockchain technology to jointly
manage the database that records Bitcoin transactions.
• That is, Bitcoin is managed by its network, and not any one central
authority.
 BlockChain in Auditing

• Blockchain represents the next step in • Accountancy profession lagged behind

accounting. other industries in digitisation.
• In Blockchain companies write to a • This is because of high regulatory
distributed ledger, cryptographically requirements for integrity of accounts
sealed. • Double Entry system allowed managers to
• Falsification, destruction or concealment is trust their own books.
impossible in Blockchain distributed • Auditors gave outsiders confidence that
ledger. accounts are ok.
• Standardisation will let Auditors verify • Audit is costly because it also takes up lot
data immediately. of accountants time.
• Fully automated audit will be possible. • Blockchain is the answer that both
• EY provides cryptocurrency wallets to all double entry system and auditors
Swiss employees, has installed Bitcoin provide.
ATM in Swiss office and accepts bitcoin as • There are private, public and consortium
payment of its consulting services. blockchains
 Closing Thoughts
• Have a passion for technology
• Be wise in its application
• Be open minded
• Be ethical
• Create an ethical environment
• Know your strengths
• Be an eternal student
• Use Youtube, Google etc – the
best Universities in the world