In an increasingly digital world, the rise of artificial intelligence is pervasive in almost all spheres of human activity. AI in simple terms is the use of intelligent machines to work and react like humans. We see many examples of AI in daily lives such as facial recognition in passport control, voice recognition in virtual assistants like Siri or Alexa, driverless cars and automated haulage systems in underground mines. The key aspect to consider is the use of AI in taking decisions in every sphere of activity where AI is applied. The crucial aspect is the quality of decision as an outcome of using AI. There are also numerous debates around whether AI will replace humans, what we can say based on the current trends in AI still is nascent and such concerns in the immediate future are unfounded. However, the key ethical question that remains is should AI be used in all spheres of human activity? How do we have controls around the usage patterns of AI? Would it infringe on personal/corporate data of people and corporations who would have no role and say in the AI decision-making process? We will consider the answers to these relevant questions in the next section.
Role Of Information Governance In AI
Every enterprise irrespective of its industry or objective generates a huge amount of data related to its operations, customers, supply chain. All this data is managed effectively using Information Governance that ensures the following:
- The data is available at the right moment in time
- The data is secure and can be trusted (of acceptable quality
- The data collected can be actioned on as it has the semantics (metadata) associated with it
These three pillars are key to enabling Artificial Intelligence applications in an organization or society at large. The key considerations around Information Governance and its role in AI are as follows -
- Compliance – enterprises leveraging AI need to be compliant with guidelines like GDPR where the rights of data subjects are pivotal. Companies running algorithms on personal data need to comply with data regulations whether it be GDPR in EU or Privacy Act in Australia. The tagging of personal data/sensitive data is managed through Information Governance controls in an enterprise and AI systems need to leverage these controls while processing sensitive data sets.
- Decision Making – AI systems running decision-making algorithms need to understand the underlying semantics of the data such as quality of data, recency of data, completeness of data to ensure quality in decision making. This is a key consideration in the usage of AI in clinical and healthcare data systems.
- Ethics – AI systems also need to consider the ethics of using AI in the context of a problem, such as can deep analysis be considered as bias against a community, can it be used to seek loopholes in the law enforcement. Here again, Information Governance plays a key role by providing the semantics of data and what kinds of acceptable processing can be done based on the compliance regulations of a given region. Government regulations around the usage of AI will develop to ensure acceptable usage patterns of AI and this can be enforced through Information Governance controls in the enterprise.
- Transparency – one of the key considerations around responsible use of AI is to develop transparent solutions to benefit humanity. For instance, the OpenAI is a nonprofit AI research company, dedicated to solving problems related to humanity. Information Governance can ensure transparency of the AI solutions by documenting the usage of data by the set of AI algorithms and ensure compliance to regulations around the acceptable usage of the given data types.
Open Data sets can also be used by AI solutions for solving problems around traffic, public utilities. Information Governance is already available in Open data sets released by governments worldwide that would enable the solutions to be transparent, compliant and ethical.
Information Governance will continue to play a pivotal role as more enterprises, public and private embrace AI technologies. The Information Governance controls will ensure that data is used for the right processing/purpose and data subject rights are not compromised. For instance, if an enterprise tries to apply AI algorithms on data which has been archived/deleted based on data subject rights, the governance controls will ensure that such processing is not allowed. For enterprises embarking on the journey of AI, must review their Information Governance controls to ensure that they are AI-ready.
Register for our upcoming events:
- Join the Grand Finale of Intel Python HackFury2: 21st Oct, Bangalore
- WEBINAR: HOW TO BEGIN A CAREER IN DATA SCIENCE | 24th Oct
- Machine Learning Developers Summit 2020: 22-23rd Jan, Bangalore | 30-31st Jan, Hyderabad
Enjoyed this story? Join our Telegram group. And be part of an engaging community.
Provide your comments below
What's Your Reaction?
Saumya Chaki works as a General Manager in the Digital Solutioning CoE at IBM by day, and an author by night. Has a Masters in Technology from IIT (ISM) Dhanbad. He has authored two books 'Enterprise Information Management in Practice' and 'A Journey Through 100 Years of Indian Cinema'. His interests include Big Data Analytics, Smarter Cities. Climate Change and Mineral Economics. His hobbies include, travel, blogging and reading. He is currently working on his third book a work of fiction.