A worker-held record of evidence, intent and terms, shared on the worker’s terms.
Fromrésumé→Toliving capability identity
The identity worth building is neither a bigger résumé nor a database about a person. It is a record the worker holds, covering what they have demonstrated, the direction they are heading and the terms on which they would work. The worker discloses it selectively, sharing only what a particular transaction needs. The technology for portable, verifiable records largely exists. What still has to be designed is control that survives bargaining pressure, because a record of availability, rate, performance and failure can serve surveillance as easily as it serves opportunity.
The person is continuous
A person may move from a university to a trainer, an assessor, an employer, a second employer and a new course while remaining the same economic participant. Institutions are episodic, and the person is the thread that connects them [CN].
The person is continuous. The institutions are interchangeable.
The concept therefore replaces the résumé with a living, permissioned and verifiable capability identity [CN]. A résumé answers what has happened to this person? The identity also answers what can this person demonstrate now, what are they becoming, and on what terms would they work?
Three layers, three defaults
The concept lists what the identity can accumulate: academic learning, micro-skills, assessments and simulations, apprenticeship history, projects, trainer attestations, demonstrated outcomes, current learning intent, availability, preferred engagement model and desired rate [CN]. These are not the same kind of information, and they shouldn’t carry the same default visibility.
| Layer | Contents | Contributed by | Default visibility (illustrative) |
|---|---|---|---|
| Evidence | Assessments, work records, attestations, outcomes | Assessors, trainers, employers, clients | Shared per transaction, at the worker’s choice |
| Trajectory | Learning intent, capability gaps, pathway | The worker and their agent | Private by default |
| Terms | Availability, engagement model, rate | The worker | Revealed only to a specific match the worker chooses to pursue |
Holding more data doesn’t make the identity valuable. It becomes different from a résumé only when its contents are verified, contextual, permissioned, portable and reusable by someone else.
The identity is not everything known about a person. It is the evidence that person chooses to use.
Holding is not the same as controlling
Technically, much of this is solved. Verifiable credentials and digital wallets let a person hold and present proof of a claim, and selective-disclosure techniques are designed to let them prove it without handing over everything behind it. States and universities are building learning-and-employment records on these foundations [1, 2]. The limit is economic rather than technical, because control is rarely tested under real bargaining pressure. In hiring, applicants hold less power than employers, and European data-protection guidance treats consent given by job applicants and employees as unlikely to be freely given [3]. A worker can hold the record and still have to hand all of it over to get the job.
So consent is necessary but not sufficient. The real protection is being able to take part without disclosing more than the transaction needs, and to decline without being penalised for it.
Consent is not control if saying no costs the opportunity.
Go deeper
A rights architecture
The identity needs enforceable rights, not just a privacy setting.
| Right | What it protects against |
|---|---|
| Selective disclosure | Handing over a full history to prove one claim |
| Purpose limitation | Evidence shared for one engagement being reused for another |
| Correction and appeal | Wrong or unfair records shaping future opportunities |
| Expiry | An old failure following someone indefinitely |
| Export and exit | Being locked into one operator |
| Opt-in, time-limited discoverability | Being permanently visible to everyone |
| Refusal without penalty | Consent that exists only on paper |
| No inferred traits | Scores or labels the worker never saw and can’t contest |
Surveillance is a documented cost
A continuously updated record of availability, desired rate, learning intent, performance, failed projects and inferred traits could become a labour-market surveillance system. The concern has evidence behind it. In a national sample of 3,508 Canadian workers, perceived workplace surveillance was associated with higher psychological distress and lower job satisfaction. The association ran through heavier job pressure, less autonomy and a sense of lost privacy, and because the study is cross-sectional, it shows association rather than cause [4]. AI systems can also infer sensitive characteristics from ordinary digital traces [3].
The design response is minimisation. Record evidence only from transactions the worker consented to. Never infer traits. Keep failures out of what employers see by default, and let the worker see everything held about them.
Rate and availability are bargaining information
Some fields in the identity are not neutral facts. Past pay is one of the most salient signals employers use to judge a candidate [5]. An employer that sees a desired rate or immediate availability may read into it the lowest wage a worker would accept, or urgency. Publishing those fields can anchor pay downward. They are better held by the worker’s agent as floors and constraints, and revealed as ranges only to a match the worker has chosen to pursue (Shift 7).
Evidence accumulates, and ages
A living identity grows richer as the same capability is observed in different contexts: an assessment, a simulation, an apprenticeship, a project with one employer and a harder project with another. That is its advantage over a static credential. It also ages. Evidence needs a date, a refresh rule and a way to lapse, or it ends up describing someone the person no longer is.
Not our invention
U.S. states are building learning-and-employment records [1]. Universities such as Arizona State give learners verifiable digital credentials in wallets, and the European Union plans digital identity wallets [2]. This series builds on that work. It adds the trajectory and terms layers, and a rights architecture designed to hold up under market pressure.
What could go wrong
- Pressure to overshare.Useful discovery may in practice demand full disclosure.
- Stale or wrong evidence.Corrections may take longer than hiring cycles.
- Creeping inference.Derived scores such as “reliability” may appear without anyone deciding to add them.
- Curation burden.Workers may not maintain the record.
- Unready employers.HR systems were built around degrees and full qualifications and may not read the record at all [6, 7].
How we’d test it
| The test | What we would measure | What would end the idea |
|---|---|---|
| Worker-control study | Use of disclosure tiers, correction, deletion, appeal and exit under real employer demand | Workers must disclose full histories to receive useful opportunities, or corrections take longer than hiring cycles |
| Identity against a rich résumé | Screening time, match quality and pay with the living identity against a résumé with references | A static résumé with references does as well or better |
| Opt-in | Share of workers who opt into discoverability, by disclosure tier | Opt-in stays too low for a working market, or harms from disclosure outweigh the gains |
How strong is the evidence?
- Established
- Portable, verifiable records are technically feasible and in use [1, 2]
- Evidence-supported
- Perceived surveillance is associated with worse worker well-being [4]; consent is weak protection in hiring [3]
- Hypothesis
- A worker-held identity with enforceable rights can be useful to employers and safe for workers
- Unknown
- Whether workers will disclose enough for useful discovery, and whether a living identity outperforms a rich résumé
Sources
- [1]National Governors Association, resource page on learning and employment recordsChecked
- [2]Jobs for the Future, paper on verifiable credential wallets in a skills-first talent marketplaceChecked (summary page)
- [3]Broecke, Artificial Intelligence and Labour Market Matching, OECD Social, Employment and Migration Working Papers (2023)Checked
- [4]Glavin, Bierman and Schieman, study of workplace surveillance and worker well-being, Social Currents (2024)Checked
- [5]Kässi and Lehdonvirta, Do Microcredentials Help New Workers Enter the Market? Evidence from an Online Labor Platform, Journal of Human Resources 59(4) (2024)Checked; the finding on pay history comes from an earlier study it summarises (Barach and Horton 2021), not re-checked
- [6]Cedefop, Microcredentials for labour market education and training: the added value for end users (2023)Checked (executive summary and chapters 1–3)
- [7]Aspen Institute, UpSkill America, report on employer perspectives on digital credentials and skills profiles (2024)Checked
- [CN]Open Commerce for Human Capital: concept document consolidated from the discussion (internal)Provided
Spotted a source we missed or a rule we misread? Tell us what we got wrong.