Bystander notice and consent have no unified protocol and depend on venue self-regulation

What it is

Museums ban photography, cafés ignore it, the office never wrote it down, and on the subway nobody can say. Capture, spatial scanning, even “looks like recording” on a headset have no cross-venue procedure for bystander notice and consent. Bystander consent today is whatever each place posts, whatever a guard stops by voice, or nobody stopping at all. A recording indicator still cannot enter the layer of “is this place even allowed.”

What is missing is not a light. It is an executable, place-varying consent path.

Why it happens

Consent on screens is usually one dialog to the user. Bystanders are not users: no account, no chance to tap agree, often no knowledge they have entered someone else’s sensor range. Moving the personal-device consent model over misses the whole unregistered crowd. Venues become the real arbiter: some ban in writing, some issue a temporary permit per event, some stay silent. Wearers read silence as permission; bystanders read it as offence; the fight starts afterwards.

With no cross-venue procedure, the same device has to re-guess local rules on every entry. The cost of a wrong guess is asymmetric: the wearer may be asked to leave, the bystander’s image may already have left the site. Venue self-regulation works when the place has staff, signage, and enforcement; streets, transit, and open campuses often have none of the three.

Studying it

Venue comparison and contextual-integrity interviews: in a gallery, an open office, a street, a hospital entrance, ask wearers and bystanders “do we ask first, whom, about what,” and observe signage and interception. Check against Nissenbaum’s contextual integrity: whether the headset breaks the expected information flow of that setting.

Venue factors: whether there is an authority, whether photography rules already exist, whether the crowd is acquainted. Scores: agreement between sides on “is consent needed,” rate of actual asking, conflict events, whether signage was seen.

Measuring only in museums that already police photography will overstate self-regulation as a general capacity.

Where it stops holding

Places the law already forbids (some courtrooms, changing rooms) do not run on self-regulation; the device should refuse capture rather than wait for a guard. In a private home the host’s consent usually covers visitor devices, but not a cloud upload the host does not know about. Rules diverge across countries and cities; “our product passed in one place” does not travel. A recording indicator lets bystanders know it is happening; it does not produce consent — knowing and allowing are two steps. When children are bystanders, venue self-regulation is almost always too thin; default deny has to sit higher.

Applying it

• Make policy configurable by venue type: capture forbidden, host permission required, verbal notice required, tracking without save. Unknown venues default to the strictest, not to fully on.
• Work with venues to turn policy into a scannable entry (signage, a geofence) so the device switches itself. Do not make the wearer guess each time.
• Split “tell bystanders” and “get permission” into two actions. With a light and no asking path, do not write “consent obtained” in the UI.
• How to check: take the device into a gallery with a photography ban, a café with no rule, and an office with a front desk. The ban should make save impossible; the office should record only under the desk’s policy; the café should be an explicit “no agreement, you decide,” not a fake authorization. Identical behaviour in all three means the venue layer does not exist.

Related

• Same group：N5.11.1 Without an externally visible recording indicator, bystanders cannot tell if they are being captured · N5.11.2 How conspicuous the device looks directly affects willingness to wear it in public · N5.11.3 Large gestures in public cause social embarrassment and suppress use · N5.11.5 Private and public places tolerate different behaviour; the same function must be designed per setting
• Nearby：N5.06 Bystanders · N2.13 Physical Safety of Input (Collisions and Bystanders)
• Search terms：bystander consent · contextual integrity · venue-specific capture policy