The AI gold rush continues into 2025 despite economic volatility. But this isn’t a race; the winner isn’t necessarily the first there. You do need to run the race to have a chance, however. In difficult times, especially for those in crisis, there is a temptation to shut down all innovation and shore up on the fundamentals. Yes, these groups must narrow their mission to focus on critical capabilities — but you can’t shut down all forms of innovation. The risk is too great. Yet innovating at a time without psychological safety can be a Herculean feat. For example, US federal agencies are undergoing massive changes. Many employees wonder every day whether their badge will work today or whether lifelong colleagues will be next to go. Budgets are also getting massively reduced with little control. Yet the mission continues, with many federal workers seeking ways to move forward with fewer resources. Automation and intelligence will be key levers for remaining resources to keep up with the mission; each are a form of innovation that must be leveraged.

Get Inspired With AI Innovation

AI tools provide many benefits at a reasonable price with high accuracy and reliability that no organization should miss. Thus far, the US federal government has focused heavily on predictive AI in its use cases (based on public-sector AI inventories), but there’s much to be gained from generative AI for employee productivity, process automation to execute more quickly, and real-time analytics to understand the current state across a massive ecosystem. Here are a couple current examples:

  • Generative AI prepares decisioning at an unprecedented speed. AI combs through such a variety of documents that human workers would just be overwhelmed by and prepares decision options other technologies are failing at. Think about finding conflicting statements in documents, identifying content changes in a sequence of document versions, or spotting gaps in documents.
  • AI plays a crucial role in enhancing transparency and accountability. By using AI to monitor government activities, agencies can ensure that they are operating within legal boundaries. Such consistent, real-time transparency helps keep public trust and confidence in government operations. Moreover, AI tools assist in detecting and preventing fraud, guiding public funds so that they are used responsibly.
  • AI automates tasks that other technologies such as RPA and DPA are simply unable to automate. These are tasks that are best described in human language as a goal rather than providing a step-by-step set of rules with exceptions. A robotic process automation (RPA) bot would immediately stop when it comes to an ambiguous point of decisioning, but generative AI can handle such ambiguities.

Get Started With AI Innovation To Overcome Resource Shortages

Now is not the time for long-term, disruptive innovations but rather tactical moves that create value fast in times of uncertainty. Leveraging AI along your innovation process increases innovation outcomes without having to reinvent the wheel: AI helps ideate based upon described challenges (in the worst case, you get a couple silly ideas that break the ice and allow humans to brainstorm with a refreshed mind and a positive attitude), identifies prior art and available solutions, designs test cases and executes testing of a proof of concept or pilot, and crafts content and messages for impacted target groups for maximum understanding and adoption of a new solution (no matter whether the solution is a piece of technology, a new procedure, or a working style). It doesn’t require huge investments in specific AI models or platforms to benefit from AI; any public large language model (LLM) will do the job.

But how do you avoid the hallucinations, compliance issues, high upfront cost, and endless rounds of discussions to remove internal resistance? Apply four recommendations as you innovate with AI. We promise that these are proven accelerators of your journey:

  • Treat AI agents like human new hires! Onboard, coach, and supervise agents and LLMs as you would with human new hires. Review their results and increase autonomy with higher trust in their ability to deliver. Ultimately, the human supervisor always stays accountable for their staff, including digital staff, but you can learn who you can trust and to what extent.
  • AI is not the answer to everything! Many things are best done by humans, some others by domain applications, and some by technologies such as QR codes or RPA. By positioning AI as one out of many technologies in your innovation toolbox, you provide a balanced approach to innovation and take out human concerns and myths in the ideation and experimentation phases.
  • Ask yourself whom you trust. Understand your level of trust in technologies, no matter if it’s AI or your system of record. This drives your innovation speed from incremental improvements toward more disruptive solutions. Ask yourself: What’s the better fit for employee- and citizen-facing processes? The system of record, the RPA bot, or the AI agent? Maybe it’s the human worker — in that case, decide for the human worker.
  • Know your process. You cannot innovate what you don’t know and don’t fully understand. Administrative processes are often complex, involving many stakeholders and requiring many execution variations. Don’t trust vendors that recommend AI to handle such complexities out of the box. You must understand your process first, as well as on an operational level. Process discovery technologies such as process and task mining along with digital adoption tools are the fastest and most accurate way to get there. Use these operational process insights to kick-start your ideation or brainstorming sessions!