Monthly Industry Trends & Insights From Colligo CEO Tim Brady
Perhaps to stay within our area of focus, my title is worth restating: AI will not solve all of your compliance and records management challenges.
Like many, I’m incredibly impressed by the rapid advancements in AI — particularly around agents, large language models (LLMs), and tools like Microsoft Copilot. But AI is not a panacea, especially when it comes to enterprise use cases and records management. Human involvement and data readiness remain the two most critical factors in determining whether AI can be effectively deployed.
In fact, many of the records managers and legal professionals I speak with are still not ready to use AI at scale — beyond basic use cases like Copilot or ChatGPT.
Where AI Falls Short in Records Management
The primary reason organizations look to AI in records management is the promise of magically applying the right retention label to the right content at the right time — in other words, auto-classification.
However, this isn’t something that can be “guessed.” It must be accurate, rules-based, and deterministic. Achieving this through rules-based automation — which already exists today — is a far better approach, as it can deliver near-100% accuracy.
AI and agents, on the other hand, tend to make estimates. That’s great for many applications, but not for environments where accuracy is non-negotiable and mistakes carry legal or regulatory consequences.
The Human + Automation Model
While knowledge workers themselves can make mistakes — such as failing to correctly designate a record — the best approach combines intelligent automation with human oversight. Let humans define the rules and boundaries, and let automation execute them.
A common (and growing) practice is to remove the end user from the process of determining a retention label altogether. Instead, solutions like Colligo can automate the classification and labelling step during filing. However, in most cases, you’d still want a user to determine whether the document is a record in the first place.
Data Privacy and Governance Considerations
With growing concerns about data security and privacy, many enterprise clients are choosing to keep their data within their own tenant. That’s not exactly compatible with feeding large amounts of sensitive information into an LLM for training.
Moreover, AI can actually amplify governance challenges if not carefully implemented.
Solving the Right Problem
When it comes to email records, our approach is to solve the actual problem — using technology and intelligent automation to deliver practical, proven value to clients, rather than adopting technology for its own sake.
This is a crucial factor in any successful technology deployment. With 80% of legal technology implementations failing to meet their original goals (according to Gartner), adopting AI simply for the sake of having AI is likely to worsen those statistics — and lead to more wasted IT investments.
The right approach is a thoughtful plan: build a solid foundation, ensure AI readiness, and choose the right technology for the problem — always keeping humans in the loop.
Disclaimer: The opinions expressed here are solely my own and do not necessarily reflect the views of Colligo. Think of them as ‘strong opinions, loosely held.’ Proceed with curiosity and a sense of humour!
