The deemed undertaking is usually discussed as a constraint on lawyers. In practice the harder problem is the expert, because an expert's working library is built out of other people's productions and the reuse happens invisibly, inside the expert's own judgment, months or years later.
All regulatory references below were checked against the Ontario e-Laws consolidations current to the e-Laws currency date of August 13, 2026. No case law is cited, and none has been verified for this note. Where the answer depends on authority rather than on the text of the rule, I say so instead of guessing.
01 The rule
In Ontario, evidence and information obtained in discovery are subject to the deemed undertaking in Rule 30.1 of the Rules of Civil Procedure, R.R.O. 1990, Reg. 194, made under the Courts of Justice Act. Rule 30.1.01(3) provides that all parties and their lawyers are deemed to undertake not to use evidence or information to which the Rule applies for any purposes other than those of the proceeding in which the evidence was obtained.
The Rule applies to evidence obtained under Rule 30 (documentary discovery), Rule 31 (examination for discovery), Rule 32 (inspection of property), Rule 33 (medical examination) and Rule 35 (examination for discovery by written questions), and to information obtained from that evidence (r. 30.1.01(1)). It does not apply to evidence or information obtained otherwise than under those rules (r. 30.1.01(2)).
The undertaking does not prohibit a use the disclosing person consents to (r. 30.1.01(4)), the use of evidence filed with the court or given or referred to during a hearing (r. 30.1.01(5)), or use to impeach the testimony of a witness in another proceeding (r. 30.1.01(6)). The court may order that the undertaking does not apply where it is satisfied that the interest of justice outweighs any prejudice to the party who disclosed the evidence (r. 30.1.01(8)).
Two features of that text carry most of the practical weight. The first is that the undertaking reaches information obtained from the evidence, not only the documents themselves. The second is that it is scoped by how the material was obtained, not by what the material is. The same document can be inside the undertaking in one matter and outside it in another, depending on the route by which it arrived.
02 Why this bites hardest on experts
A schedule expert accumulates material at a rate no other participant does. On a single delay file the production can include the native schedule files for every update, the daily reports, the inspection and survey records, the meeting minutes, the change and request for information logs, the correspondence, and often the labour and equipment records behind a productivity claim. All of it arrives under Rule 30 or comes out of Rule 31, and all of it is far more granular than the pleadings.
That material is also, from the expert's point of view, the raw feedstock of expertise. Productivity rates, achievable installation durations, realistic procurement lead times, how a particular class of schedule behaves under update: an analyst gets better at all of it by reading other people's records. The temptation to carry it forward is not a temptation to cheat. It feels like getting better at the job.
The realistic failure mode is not an expert emailing a produced document to a new client. It is an expert who, in matter B, uses a benchmark, a rate, a lead time, or a characterisation of a party's scheduling practice that exists in the expert's head only because of what was produced in matter A, and who cannot now say where it came from. That is a use of information obtained from evidence to which the Rule applies, and it is invisible to everyone including the expert, until the expert is asked on cross-examination for the source of the figure and has no answer.
03 Three things that get run together
It helps to separate three categories, because they are treated differently and only the third is genuinely unsettled.
The document
A document produced under Rule 30, or an answer given on an examination for discovery under Rule 31, cannot be used for the purposes of another proceeding. This is the easy case and it is absolute in practice. It covers the native schedule file as much as the printed Gantt chart, and it covers the working copy the expert made as much as the copy as produced.
The information taken from the document
Rule 30.1.01(1) extends the Rule to information obtained from that evidence. So the number transcribed out of a produced daily report is in the same position as the report. So is a derived figure: an installation rate calculated from produced quantity and labour records, a float profile extracted from a produced schedule, a summary table built from produced correspondence. The transformation does not launder the source.
Generalised professional learning
Everything an analyst reads changes their judgment slightly. Nobody thinks the Rule requires an expert to forget how to do the work. The line between information obtained from produced evidence and general expertise developed while working with it is not drawn by the text of the Rule, and I am not going to pretend the text answers it. Nor am I going to cite authority on it, because none has been verified for this note.
Draw the line at traceability. If a statement in a report in matter B can be traced to a specific document produced in matter A, it does not go in, whatever form it has been transformed into. That includes a rate, a duration, a lead time, a quantity, a productivity factor, and any characterisation of how a party or a project behaved. What may go in is method, technique, and the judgment that comes from having done the work, expressed without carrying the underlying data forward. This test is stricter than the outer limit of the Rule may turn out to be. It is also the only version of the test that can actually be applied by an expert working alone at nine at night, and it is the version that survives cross-examination without argument.
04 Subrule (2) is the most useful provision, and it runs on provenance
The Rule does not apply to evidence or information obtained otherwise than under those rules (r. 30.1.01(2)). On a construction file that carve-out is wide. Much of what a schedule expert works with came from the client's own project records: their own schedule files, their own daily reports, their own correspondence, their own progress photographs. That material is not inside the undertaking merely because the same documents were also produced in the litigation.
The catch is proof. Subrule (2) is an assertion about the route by which a document arrived, and an expert who cannot show the route cannot rely on the subrule. On a large file the same document exists in three places: the client's transfer, the production set, and the expert's working folder, with different file names and often different dates. Once those have been mixed, the provenance answer is guesswork.
Keep a provenance ledger per matter, one row per document or transfer: what it is, who supplied it, the date received, the route (client transfer, production, undertaking answer, examination exhibit, public source), and where it now sits. Record it at intake, not at report stage. It costs minutes per transfer, it is the only way to answer a provenance question honestly a year later, and it converts subrule (2) from an argument into a record.
05 The exceptions, and what they really allow
The three exceptions in subrules (4), (5) and (6), plus the court's power under subrule (8), are narrower than they read when someone is looking for a way to reuse material.
- Consent. The undertaking does not prohibit a use the disclosing person consents to (r. 30.1.01(4)). This is the clean route and it is the one people forget. If there is a good reason to reuse material, ask. A refusal costs nothing beyond the time it took to ask, and an unrecorded assumption of consent costs everything.
- Filed or used at a hearing. The undertaking does not prohibit the use of evidence filed with the court or given or referred to during a hearing (r. 30.1.01(5)). This is the legitimate source of published examples and teaching material. It requires the expert to know, document by document, what was actually filed or referred to, which is a narrower set than what was in the brief.
- Impeachment. The undertaking does not prohibit use to impeach the testimony of a witness in another proceeding (r. 30.1.01(6)). This is counsel's tool, and it does not license an expert to build an opinion in matter B out of matter A material on the theory that the same witness might appear.
- Court order. The court may order that the undertaking does not apply where it is satisfied that the interest of justice outweighs any prejudice to the party who disclosed the evidence (r. 30.1.01(8)). This exists, and it is not a plan. If reuse matters enough, it is a motion, brought before the reuse rather than after it is discovered.
06 Controls that actually work
These are the controls I would expect to see in place, and the ones counsel should ask about before retaining a schedule expert. They are equally a checklist for the expert.
Matter siloing at the file system, not at the level of intention
One storage root per matter, no shared working folders, no cross-matter search index spanning matters. An expert who searches their whole archive for a productivity rate has already crossed the line before deciding anything.
No cross-matter comparables database
The idea of a private library of durations, rates and productivity factors built up over past engagements is attractive, and it is the single most direct route into breach if any of it came out of a production. If such a library exists it has to be sourced exclusively from material outside the undertaking, and the source of every entry has to be recorded, entry by entry. My view is that the recordkeeping burden is larger than the value, and that published sources and the current project's own records are the better foundation.
Keep produced native schedule files out of the shared working environment
This one is specific to scheduling work and it is routinely overlooked. Importing a produced schedule file into a live scheduling database does not import a project in isolation. Calendars, resource dictionaries, code structures and similar global objects come with it and become available to every other project in that database, and they will appear in exports from unrelated matters. Import produced schedules into a dedicated database or a standalone instance for that matter alone, export nothing from it into a shared environment, and keep the original file untouched alongside a hash so the working copy can always be tied back to what was produced.
An engagement letter that names the obligation
The retainer should record that the expert is bound by the undertaking through counsel, that material received for the matter will be used only for that matter, how the material will be stored, who else will see it, and what happens to it at the end. A staff or subcontracted analyst is bound by the same terms in writing before touching the file.
A disposal step that is actually performed
Return or destroy on the timetable counsel sets, confirm it in writing, and record what was retained and why. Some retention is legitimate: the expert's own working papers may be needed if the opinion is later challenged. That is a reason to keep a defined set, not a reason to keep everything forever in a folder that will be searched by someone else in five years.
A sole practitioner or small practice cannot rely on institutional walls, conflicts software, or a records department, so the discipline has to be structural: file layout, a provenance ledger, a separate scheduling database per matter, and a disposal step. Built that way, the controls are cheap. Bolted on after a problem appears, they are not available at all, because the mixing has already happened.
07 Cross-examination, and writing about past matters
Two situations put an expert on the spot and both are foreseeable.
The first is the question about prior experience. Asked whether they have seen a particular pattern before, an expert can properly speak to their general experience and to method. They cannot describe what was in documents produced in another matter, and they should not be drawn into it by a question that sounds like an invitation to demonstrate depth. Counsel should raise this with the expert before the examination rather than watch it happen, and the expert should have a prepared, non-evasive form of words: what they can speak to, and why they are not going further.
The second is publication. Writing, speaking and teaching are how a schedule expert becomes known, and past matters are the only interesting material. The route through subrule (5) is real but narrow: evidence filed with the court, or given or referred to during a hearing. Anything else needs the disclosing person's consent under subrule (4).
Write from the filed record or with recorded consent, and get the client's approval for the specific use even where the material is technically outside the undertaking. Anonymising a project does not answer the Rule, because the Rule is about the use of the evidence and information rather than about naming. It also does not work in practice: a construction project is identifiable from its dates, scale and sector to anyone in the industry who was near it.
08 What is uncertain, and what I have not verified
- Whether, and how, the undertaking binds a non-party expert directly is not resolved here. The Rule speaks of all parties and their lawyers. The practical answer is to bind the expert contractually through the retainer and to have counsel supervise, which makes the theoretical question academic on a well run file. It is not academic if it ever has to be litigated, and I do not answer it.
- Where generalised professional learning ends and information obtained from produced evidence begins is not answered by the text of the Rule. Section 03 gives the conservative line I use. It is a working rule, not a statement of the law.
- No case law is cited, so nothing here reflects how courts have applied the Rule in particular circumstances, including on motions under subrule (8).
- Rule 30.1 is a court rule. It does not govern an arbitration, and confidentiality in arbitration comes from the agreement, the applicable statute and any institutional rules. That machinery is not addressed here, and it should be checked before assuming either that an arbitral record is freely reusable or that it is protected.
- Obligations arising from other sources, including privacy legislation, confidentiality clauses in the underlying construction contracts, and professional obligations owed by counsel, are outside this note and may be more restrictive than the Rule.
09 Practical points
- Assume everything received on a matter is inside the undertaking until the provenance record shows otherwise.
- Start the provenance ledger at first transfer, not at report stage. What it is, who sent it, when, by what route, where it sits now.
- One storage root per matter. No shared working folder, no archive-wide search.
- Import produced schedule files into a dedicated database or standalone instance, never into a shared scheduling environment.
- Hash the file as produced, work on a copy, and keep the two associated.
- Do not build a cross-matter library of rates, durations or productivity factors.
- If a figure in a report cannot be traced to a source that is properly available in this matter, it does not go in the report.
- Ask for consent under subrule (4) when reuse would genuinely help. Record the answer either way.
- For anything published, work from what was filed or referred to at a hearing, and get the client's approval for that specific use.
- Put the undertaking, storage, access and disposal terms in the engagement letter, and bind any assisting analyst in writing.
- Brief the expert before examination on how to answer questions about prior matters.
- Perform the disposal step, confirm it in writing, and record what was retained and why.