Assessing What Is Known
Grace, Jack, and Mia, the company’s legal adviser, review the investigation’s early findings. They distinguish confirmed facts from open questions, including what data was accessed and whether it was taken.
Dialogue12

Let’s separate what we can substantiate from what’s still inference. Jack, what did the logs establish overnight, and what remains uncertain?
We confirmed that an external account accessed the customer database on Tuesday. The account had valid credentials, but we haven’t established how they were obtained or whether the access was authorized.
When you say “accessed,” do we know which records were actually viewed, or only that the account could reach the database? That distinction will matter in any notice.
The logs show queries against two tables containing names, email addresses, and hashed passwords. We can’t yet tell whether the results were displayed in full or exported from the system.
So the presence of a query is confirmed, but the extent of exposure isn’t. Do we have evidence that payment details or account recovery information were involved?
Not so far. The payment environment is segmented, and its access logs show no unusual activity. That’s reassuring, but I wouldn’t call it conclusive until we validate the logging coverage.
And do we know whether any data left our systems? I want to avoid using “stolen” unless we can support that specific claim.
We found a large outbound transfer during the same window, but it may include routine backup traffic. The destination is under review, and we’re comparing the file size with normal patterns.
Then our working account should say an account accessed certain database tables, while the possibility of data extraction remains under investigation. Is that accurate enough?
Yes, provided we state what’s known without implying that every affected record was viewed. We should also document the basis for each claim and update it as evidence changes.
I’ll preserve the relevant logs and snapshots, and keep the investigation team from altering the affected systems until we’ve captured what we need. We’ll also verify whether the account is disabled everywhere.
Good. Let’s maintain a fact-and-question tracker: confirmed access, possible exposure, no confirmed extraction, and payment data not implicated so far. We’ll revisit each point when the evidence comes in.
Vocabulary7
- substantiate
To support a statement with evidence. Here, Grace wants to distinguish claims the investigation can prove from those that are still assumptions.
- inference
A conclusion drawn from available information rather than directly confirmed evidence. It is useful when carefully describing what may have happened.
- valid credentials
Login details that the system accepted as legitimate. The phrase confirms how the account gained access without establishing who obtained the credentials or whether the access was authorized.
- logging coverage
The extent to which system activity is recorded in logs. This phrase helps explain why a lack of unusual activity may not be conclusive.
- under review
Being examined or investigated before a conclusion is reached. Use it to describe an unresolved issue without implying an outcome.
- working account
A provisional explanation of events based on what is currently known. It may change as new evidence becomes available.
- fact-and-question tracker
A record that separates established facts from issues still being investigated. It helps teams update their understanding without treating open questions as settled.
Key Sentences7
- 1
Let’s separate what we can substantiate from what’s still inference.
This is a useful way to begin a careful discussion when some claims are supported by evidence and others are not. You can adapt it to separate confirmed information from assumptions in a report or meeting.
- 2
We confirmed that an external account accessed the customer database on Tuesday.
This sentence states a specific finding while identifying who or what was involved and when. Reuse the structure “We confirmed that…” to report verified information.
- 3
We can’t yet tell whether the results were displayed in full or exported from the system.
This clearly identifies an unresolved question without suggesting that either possibility is proven. Use “We can’t yet tell whether…” to describe uncertainty precisely.
- 4
That’s reassuring, but I wouldn’t call it conclusive until we validate the logging coverage.
This balances a positive indication with a caution about the limits of the evidence. The phrase “I wouldn’t call it conclusive until…” is useful when a finding needs further verification.
- 5
I want to avoid using “stolen” unless we can support that specific claim.
This expresses a careful approach to language when a claim could have legal or reputational consequences. You can use “avoid calling it [claim] unless…” to emphasize that wording should match the evidence.
- 6
We should also document the basis for each claim and update it as evidence changes.
This sentence recommends recording how claims are supported and revising them when new information emerges. It works well in professional settings where accuracy and accountability matter.
- 7
Let’s maintain a fact-and-question tracker: confirmed access, possible exposure, no confirmed extraction, and payment data not implicated so far.
This summarizes what is known and unknown in a concise, organized format. You can reuse “Let’s maintain a [type of] tracker” to propose a shared record for monitoring an investigation.