SecurityBrief Asia - Technology news for CISOs & cybersecurity decision-makers
Story image
Factors to consider before undertaking an AI deployment
Tue, 12th Mar 2024

With the pace of artificial intelligence (AI) development showing no sign of slowing, attention is now moving to the benefits and opportunities it can deliver.

Since ChatGPT first demonstrated the capabilities of generative AI in late 2022, business leaders have been evaluating how the technology can be put to work. They’re assessing how AI-powered tools can streamline workflows, improve productivity, and reduce costs.

At the same time, CIOs within many organisations are also coming to terms with the complexities that the introduction of AI tools can create. Because the tools can work with data drawn from a variety of departments and teams, CIOs are realising they will need to work closely with other business leaders during deployments.

AI tools also have the potential to cause fundamental shifts in workflows and patterns. Tasks that traditionally have required human intervention can increasingly be automated and streamlined. Close consultation with staff at all levels will be required so that they understand the effect that such changes will have.

Taking a strategic approach
For the deployment of AI-powered tools to be effective, it’s necessary for organisations to take a strategic approach to their introduction. Rather than focusing on what they will deliver in isolation, managers need to consider their impact on wider operations and explore broader potential synergies.

Many organisations are also finding it beneficial, to begin with a pilot project to build up an understanding of the capabilities of the tools and how they can best deliver business value. Once this has been completed and assessed, a strategy can be developed for wider deployments across the broader organisation.

This explorative approach will be particularly important in areas where sensitive data is held and processed. Examples include the legal and medical sectors where the security of client and patient records is paramount.

Dealing with ever-growing volumes of data, especially unstructured and poorly organised data, will also become challenging for many organisations. New processes will be required for its storage and management to ensure AI tools can perform as efficiently as possible.

Time will also be required so that ongoing advances in AI technology can be carefully evaluated and assessed. New capabilities need to be better understood as they are introduced into the workplace, noting the risks and opportunities they present. 

This is particularly important when it comes to testing features such as machine-learning algorithms and neural networks. It will also be important that organisations can respond to demands by regulators for evidence of how particular results were created.

The human element
Business leaders also need to be cognisant of the impact that AI-powered tools will have on their employees. Some may be excited by the prospect, while others could become fearful that their roles will no longer be relevant.

Time must be taken to explain exactly how the tools will be used and the benefits they will deliver. Staff need to be assured that they’ll augment their roles rather than replace them. In many cases, the tools will allow staff to give up repetitive tasks and focus their time on more value-adding activities.

Staff will also need to understand the importance of having human oversight of AI tools and the processes they undertake (where this is practical!). Outputs cannot simply be taken at face value in all cases. Instead, they need to be assessed against what was expected to ensure the tools are performing in the correct manner.

Effective governance
A further area that will require ongoing monitoring and review once AI tools are in place within an organisation is governance. Senior leaders need to be sure that data used by the tools is protected and remains secure.

It’s also vital that organisations adhere to all compliance regulations relating to data retention and usage. Acknowledgement of privacy requirements must occur at all times.

AI-powered tools will have a wide impact on organisations across all sectors of business in the years ahead. By taking a thoughtful and methodical approach to their deployment and usage, businesses will be able to enjoy the significant benefits that they will deliver.