Do we need to replace our existing systems?
No. AI should build on what works. Replacing a system that does its job in order to ‘do AI’ is one of the surest ways to waste time and lose the teams’ trust.
Updated on 24 October 2026
What is going on.
The arrival of AI sometimes comes with a temptation: to use the opportunity to overhaul everything, replace the ageing software package, the in-house tool, the spreadsheets. Some suppliers encourage this, offering platforms that would do everything at once.
These replacements are long, expensive and risky. They tie up teams for months, disrupt what was working and push back the gains expected from AI by as much. Yet AI can usually build on what exists: read its data, feed its screens, slot in between two steps. You only replace what really stops you moving forward.
What we observe.
- An old system is not necessarily a hindrance.If it does its job and the teams know it well, it is often an asset.
- Big replacements absorb all the energy.Meanwhile, no new use reaches the teams.
- The real blockages are often localised.Inaccessible data, a missing interface, a file that cannot be shared.
What you can do right now.
- For each use you are considering, identify what it needs from your current systems.
- Pinpoint exactly what is blocking, and first try to work around it or fix it.
- Be wary of an AI project that starts by replacing software.