When a work tool stops behaving, the message to IT often looks like this. My laptop is not working. Please fix it urgently. It sounds clear to the person sending it, but it leaves the person receiving it with almost no useful starting point.
AI-powered help desks are becoming more capable. On September 15, 2026, Fujitsu announced Aurora, a multilingual voice-led workplace support agent designed to help resolve IT problems across channels. The company says it supports interactions in more than 50 languages and can coordinate diagnosis and follow-up. These are vendor claims about the announced service, not a promise that every workplace has it today.
The practical lesson for employees is less about the new interface and more about the information going into it. Whether a human or AI assistant responds, a precise first message can reduce unnecessary back-and-forth.
Five details that make a support request useful
Use this structure
- What you were trying to do State the task, not just the problem.
- What happened instead Include the exact error message if one appeared.
- Where it happened Name the application, device and relevant browser or operating system.
- When it started Say whether it is new, intermittent or happening after a change.
- What you already tried Include safe steps already completed and what happened after each.
Add the business impact when it matters. Saying that a customer call begins in 20 minutes gives support staff useful context. Calling everything critical does not.
The best support request tells someone what failed and gives them enough context to reproduce it.
Turn a vague request into one that can be investigated
Here is a fictional example. A salesperson writes that the CRM is down. That could mean the entire service is unavailable, one account will not open or a report is taking longer than usual.
A better message says that the salesperson was exporting a weekly pipeline report from the CRM in Chrome on a company laptop. The export button returned an access denied message at 9.15 am. The same account could still open customer records. Logging out and back in did not help. The report is needed for a 10 am meeting.
That description does not solve the issue automatically. It gives the support team a narrower starting point, a reproducible symptom and a reason to prioritise it.
Try this now before sending your next ticket
Copy-paste support request template
I was trying to [task] in [application] on [device or browser].
Instead, [what happened and exact error wording].
This started [time or date] and happens [every time or sometimes].
I already tried [safe steps] and the result was [outcome].
The impact is [who or what is blocked and relevant deadline].
I can share a screenshot with sensitive information removed if needed.
Use the same structure when talking to an AI support assistant. If it suggests a fix, check whether the instruction applies to your organisation and device before following it. Do not grant new permissions, disable security settings or share passwords just because a chatbot asks.
When a support bot is not enough
Some requests need a person, an administrator or a formal approval. Examples include a suspected phishing message, missing customer data, unusual account access, a security warning or a change that affects many colleagues. Use your employer’s established escalation channel for these situations.
For an ordinary issue, ask the assistant to repeat the problem it understands, list the next safe checks and say when escalation is needed. If the conversation becomes circular, send the five-detail summary to a human rather than starting from zero.
A quick check before you submit
- Does the message name the task and the affected tool?
- Is the error wording exact rather than paraphrased?
- Can another person tell when the issue began?
- Have you described the impact without exaggerating it?
- Have you removed passwords, access tokens and unnecessary personal data?
As AI becomes a more common first stop for workplace support, this small habit remains useful. Good input helps a support system understand the problem, and a clear record makes it easier for the next person to continue the investigation.
Sources
Fujitsu – Aurora workplace service desk announcement – September 15, 2026
