Why a failure to plan with AI, is a plan to fail

Artificial intelligence has been recognised globally as a key to unlocking impressive efficiencies and operational gains, both by organisations and individuals. In the public sector, AI can help transform public services. It can streamline processes and provide employees with more agency, freeing people up and saving team time. This means employees can use their time much more effectively. In fact, a recent trial of Microsoft’s Copilot across Whitehall, involving 20,000 civil servants over three months, found that AI can save officials an average of 30-minutes a day, amounting to nearly two weeks over an entire year. In such an important industry, such increases in productivity can have tangible impacts in driving growth and, ultimately, in enhancing the lives of citizens.

With such gains on offer, the UK government has made it clear how crucial it views AI readiness. The launch of its dedicated AI Opportunities Action plan at the start of the year stands out. The plan recognises how the public sector stands to benefit from using AI to remove repetitive, monotonous and time-consuming administrative tasks. A suite of AI solutions, labelled ‘Humphrey’, has been designed by the government to speed up work for civil servants. This is cutting the time spent on admin and reducing the money spent on contractors.

The UK government has identified data as a strategic asset; recognising that the foundation of AI is data. Data is the critical enabler of transformative technologies. It is how these advanced systems learn. It is also what they are trained on. Yet, the public sector will face many challenges when preparing data to fuel any AI ambitions due to legacy foundations and data siloes. This makes it extremely difficult for data, and by extension AI, to be properly maximised.

At NetApp, we know how AI needs reliable, secure data to fuel successful outputs. Unmanaged, disorganised and siloed data could limit the productive potential and cost cutting impact the UK government is keen to see from investments in technology and AI. Before the public sector becomes fully AI-ready, it must first become data-ready. However, several challenges concerning quality and security facing the sector’s data pool must be addressed first.

To fully understand the extent of these challenges, firms must step back and evaluate their entire ecosystem. Why? Well, organisations will find it incredibly difficult to adopt AI with scattered and misplaced data across various sources and systems. Processes in the public sector are also often siloed, which has hampered technological adoption for several years. If AI is to succeed, it needs commonality and shared accessible data.

The first step of any project should be to identify and connect data, well before anything is processed. By understanding the full data landscape, organisations can better clean and organise data, and therefore ensure AI effectiveness. As part of the AI Opportunities Action plan, the UK government has committed to ”responsibly, securely and ethically unlock[ing] the value of public sector data assets” as part of a National Data Library.

On a sector level, this National Data Library will bring together widely available public data from different sources with the aim of training AI. It is essential to the sector accessing actionable insights from AI use. However, it is about both quality and quantity when it comes to data, and for the UK government’s AI ambitions to succeed, a robust data management strategy needs to be implemented as a priority.

Nothing underpins the urgency of this quite like the spate of cyberattacks we’ve seen recently in the UK on national retailers and public bodies. This highlights the importance of protecting data. MP’s fear that public sector departments are not prepared for attacks, worrying that the UK is not equipped with effective recovery strategies.

As public bodies ramp up data sharing and data collection in a bid to fuel AI and analytics tools, it is imperative that this data is protected properly. Not only will it help avoid possible cases of data poisoning, but it also prevents the extreme case of irrecoverable sensitive data loss.

The signal for AI readiness from the UK government relates to a need for and recognition that data should be driving activity across the sector. It is not just processes that are siloed, it is crucial data that is unable to be safely shared or utilised across departments.

On top of this, data challenges cannot be overlooked. The UK government risks promoting and implementing limited AI models, if the data they’re relying on is unprepared and unprotected. What should be “informed” data driven decision making risks becoming “somewhat accurate” data driven decision making. The right approach to data will either make AI a success, or will quite literally break it.

Central Government Programme activities

The techUK Central Government Programme provides a forum for government to engage with tech suppliers. We advocate for the govtech sector, evangelise tech as a solution to public sector challenges, facilitate market engagement, and help make the public sector an easier market to operate in. Visit the programme page here.