AI is arriving in the spreadsheet. Is the spreadsheet still the right system?

AI assistants are moving into the spreadsheet itself. When Anthropic released Claude Opus 4.6 on February 5, 2026, it listed an upgraded Claude in Excel and a research-preview Claude in PowerPoint among the launch features ^[1]. For a business that runs on spreadsheets, that is useful news, and it also raises a design question.

What AI does well in a spreadsheet

Anthropic says Claude in Excel now plans before acting and can make multi-step changes in one pass ^[1]. In our view, that kind of help suits the tedious parts of a sheet, such as drafting a first version of a model, restructuring tabs and cleaning inconsistent columns. We have not measured this, and the evidence in this brief is limited to the vendor's description of the feature. If a spreadsheet is a calculation tool, an assistant plausibly makes it faster to build.

What AI does not change

A spreadsheet that has become a system of record has problems unrelated to how quickly it is edited. Microsoft documents a worksheet limit of 1,048,576 rows by 16,384 columns, and, for the legacy shared-workbook feature, a limit of 256 users who can open the file at the same time ^[2]. Those limits are generous for analysis, but a process that tracks jobs, approvals or customer records is asking for things a sheet does not provide natively: enforced data types, per-role permissions, an audit trail of who changed what, and relationships between records that cannot drift apart.

Signs a process has outgrown the sheet

• Several people keep their own copies, and reconciling them is a recurring task.

• A single person is the only one who understands how it works.

• The sheet gates real actions, such as sending invoices or releasing orders.

• You have added colour codes or hidden columns to simulate statuses and permissions.

What an internal tool changes

A purpose-built internal tool gives that process a defined data model, role-based views and an admin dashboard, so the people who work in it see only what they need. It is the territory covered by Sinthetix's custom software work: internal tools shaped around a team's exact workflow, admin and operations dashboards, and database design that includes a migration strategy for the data you already have. The process runs through discovery, architecture, iterative builds and a documented handoff.

The two approaches combine well. An AI assistant can help your team clean up and map the existing spreadsheet during discovery, and the new tool can then export back to a sheet wherever analysts still prefer to work.

For most businesses, the practical step is to list the three spreadsheets people would be most worried about losing, check each against the signs above, and move only the ones that act as systems of record.

Talk to us: sales@sinthetix.com