Iron Mountain Ltd: Emerging AI technologies shaping the future of digital transformation #techUKDigitalPS

Enabling better access to digital technologies across the public sector is a top priority for the UK government. Last year, the government released £23 million to create 2,000 scholarship programs in the rapidly evolving fields of data science and artificial intelligence. The hope is that such programmes will help close the skills gaps across both the public and private sectors.

Nonetheless, the volume and complexity of data continue to grow as public sector companies digitise their records, some of which span back decades. Both digital transformation and data management are vital to delivering high-quality services to citizens, mitigating security risks, and ensuring compliance with current and future legislation.

Artificial intelligence (AI) shows great promise in helping organisations, such as the NHS, scale their data operations. While AI itself is nothing new, it’s easily the fastest developing field in technology. In this article, we’d like to share how some of the key emerging trends in AI have profound implications for digital transformation.

Opening the black box with explainable AI

One of the biggest challenges of adopting AI in the public sector is that the technology often operates in what we call a black box. AI solutions can produce results at a level of complexity that only a computer can comprehend. In other cases, vendors of AI solutions may not disclose how their models work or where they get their training data from. In either case, humans don’t know how the system arrived at its conclusions.

The result is a lack of transparency and accountability, which can have potentially disastrous consequences in the public sector. Unsurprisingly, many critics have raised concerns about the use of fully automated decision-making in sensitive areas like law enforcement, education,  and healthcare. One of the more poignant examples was the use of algorithmic grading in the UK during the pandemic, which was quickly abandoned following a severe lack of alignment between AI and the expectations of teachers.

Explainable AI seeks to address these concerns by making AI development more transparent, understandable, and fair with the ultimate aim of leveraging it to augment, rather than replace, human decision-making . A rapidly emerging focus area in AI model development, explainable AI encompasses the tools and frameworks needed to help people understand how an algorithm arrived at its answers. In use cases like digital transformation or data management, explainable AI can help improve accuracy and consistency when digitising records from multiple formats, including non-standard or obsolete ones. Other benefits include categorising data and maintaining audit trails that track the entire decision-making process.

Preserving privacy with federated learning

Adopting the principles of explainable AI can help alleviate these concerns, but it’s not enough by itself. Federated learning takes AI model development an important step further by having multiple organisations involved in training machine learning models, albeit without having to share their sensitive data. There’s potential for federating learning to incorporate blockchain networks as well, to further enhance digital security and integrity with an immutable record of the training process. Blockchain networks offer decentralised storage without the need for any centralised authority. They’re ideally suited to AI model development in highly sensitive use cases, such as healthcare and finance, where security and privacy are paramount.

While explainable AI and federated learning aren’t especially new concepts, they’re becoming all the more important with the meteoric rise of new AI models, such as generative AI or deep learning networks. Together, these emerging trends offer transparency, accountability, and privacy to help deliver more effective services to citizens. They will continue to prove vital in unlocking the full potential of AI to build a better digital future for all. 

This article was written by Aaron Kalvani, Senior Solution Engineer at Iron Mountain EMEA.

 Aaron Kalvani is a global leadership expert on 'Artificial Intelligence coupled with Information Lifecycle Management'.  Working with the United Kingdom' Central Government 25 years, he continuously stays way ahead of the technology curve and evangelises his deep knowledge in high-energy keynotes presentations that challenge audiences to leverage their focus and pay attention to what matters most in building a reliable, scalable, future proof and sustainable integrated smart architectural solution and platform. To learn more about Aaron, please visit his LinkedIn.

To learn more about Iron Mountain EMEA, please visit their LinkedIn and Twitter.

To read more from #techUKDigitalPS Week, check out our landing page here.

You can also follow the campaign on techUK's Twitter and LinkedIn - #techUKDigitalPS.