Shrinking budgets, rising citizen expectations and increasingly sophisticated fraud schemes have stretched resources thin. Meanwhile, outdated systems and manual processes make it harder for frontline workers to keep pace.
According to a global study commissioned by SAS, a striking 85% of policymakers and public sector leaders cite the fight against fraud, waste and abuse (FWA) as one of their top five priorities – and for good reason. It’s estimated that around 16% of public budgets are lost to FWA, undermining not only financial efficiency but also public trust.
Still, there’s reason for optimism: 100% of surveyed civil servants report that the use of data and AI has already improved their productivity in tackling these irregularities. For teams under constant pressure and working with limited resources, that’s a welcome shift. And this is just the beginning.
As a strategic advisor in digital transformation working with public sector teams across European countries, I see fraud prevention not just as a control mechanism but as a powerful catalyst for broader work transformation. When done well, anti-fraud strategies are not solely about compliance – they’re about freeing up time, sharpening focus and creating greater societal value.
Allow me to explore this further in seven concise and practical points.
1. Reframe AI as a productivity partner – not just a fraud detector
AI is often introduced as a surveillance or compliance tool. But in practice, its role is far broader and more empowering. In the SAS study, 57% of respondents cited “working more efficiently” as the main benefit of AI, even ahead of fraud detection.
Rather than replacing human capabilities, AI can:
- Pre-screen low-risk claims so analysts can focus on complex cases.
- Summarize large case files to accelerate decision-making.
- Flag suspicious transactions in real time, helping reduce backlogs.
The result? Time saved, reduced burnout, and a team that spends less time on irrelevant details and more on what truly matters.
2. Automate the tedious to make room for the complex
Fraud investigations often require combining data from different systems, verifying identities and tracing transactions over time. These are typically repetitive, rule-based tasks – ideal for automation.
A European social security agency increased its productivity by using AI for automated identity and eligibility checks. This gave staff more time and space to focus on complex cases.
Key takeaway: Automate the “administrative work” so professionals can apply their expertise where it has the most impact.
3. Equip teams with the right skills, not just the right tools
One of the biggest obstacles in fraud prevention? A shortage of analytical skills – cited by 46% of respondents. Even the best technology won’t deliver productivity gains if teams don’t know how to use it effectively.
What can organizations do?
- Ensure all employees have a basic understanding of data analysis. (Data literacy is also embedded in the EU AI Act.)
- Encourage collaboration between analysts and fraud experts to interpret insights accurately.
- Promote explainability in AI systems so users understand why something is flagged as suspicious.
Insight: AI should be a trusted tool, not a black box.
4. Create “quick win” use cases to demonstrate value
Implementing AI across an entire department at once is often too ambitious. That’s why SAS strongly recommends starting with manageable, high-impact projects, and I fully agree.
One government agency improved its detection of identity fraud using AI-driven alerts and a smart triage system. Within a few months, case resolution times improved noticeably.
From a productivity perspective, this had a dual effect:
- Less time lost on false positives.
- Greater motivation comes from seeing tangible results.
Tip: Start small, demonstrate value, and scale with confidence.
5. Use cross-agency collaboration to reduce duplicated effort
Fraud doesn’t stop at organizational boundaries, but data often does. Only 7% of respondents said they regularly collaborate with other government agencies on fraud prevention. That’s a huge missed opportunity.
While legal constraints exist, sharing data and insights between federal, regional and local tax authorities, customs, social services, procurement, inspection bodies, and even the financial sector can help teams:
- Detect early warning signs (e.g., inconsistencies across systems).
- Avoid duplicate investigations.
- Learn from each other’s tools and methods.
Point: Collaboration isn’t a burden, it’s a chance to work smarter and faster.
6. Don’t overlook unstructured data because it’s a hidden productivity goldmine
Most fraud detection systems rely on structured data: names, dates, amounts. Yet only 28% of organizations use unstructured data despite its immense value.
Think of attachments in tax filings or import documents, various emails, handwritten forms, free-text XML fields, transcripts of customer conversations or even social media comments. Traditionally hard to process, but AI makes the difference here.
Thanks to techniques like natural language processing and digital twins, government agencies can extract valuable insights from vast amounts of unstructured data.
In practice: Less time spent manually compiling case files, more time focused on core risks.
7. Make productivity gains visible to the wider organization
In the public sector, productivity is often seen as an internal benefit. But if fraud prevention leads to faster service, shorter wait times and better outcomes – share that story!
This helps to:
- Build internal support.
- Strengthen public trust.
- Justify further investments in tools and training.
In short: Every efficiency gain is a chance to highlight the importance and impact of your work.
Final thought: Fraud is a challenge, but also a lever
Fraud, waste, and abuse will never fully disappear. But governments today have stronger tools than ever to fight back with more control, clarity, speed and strategy.
The real opportunity? Use this challenge as a springboard to fundamentally rethink public sector work – not just safer but smarter, more efficient and with greater societal impact.
The ultimate goal isn’t just less fraud, but a stronger, more capable government that builds trust in and among people and contributes to a fair, caring and collaborative society.
