Skip to content

PAN Lab levers

Lever

Keep the door open after hours

The AI assistant answers when the office is closed. Nobody waits until Monday for help that the assistant could give on Saturday. Nobody is on shift, either, to check what the assistant says in those hours.

What it is

A service that closes at the end of the working day is closed to everyone who works the same hours. After-hours coverage lets the assistant answer questions, take requests, and give routine help at night and on weekends. The survey experiment cited below found that people's intention to use a government chatbot tracked how useful they found it and how much they trusted the technology. Their privacy concerns attached to the chatbot's service uses.

What it pushes on in the Lab

In the Lab, this lever strengthens the pathway by which people receive the automated system's output, because the service now answers people it could not answer before. It also weakens the output checks that screen the model's output before people act, because nobody is on shift to challenge it. The same pathway carries the assistant's errors too, so when you pull this lever, a wrong answer reaches people as readily as a right one.

amplifiedFailures adopted by people or agents
dampenedOutput checks screen model output before people act

You can also pull this lever at a strong tier, which costs more. At the strong tier, each effect below the Lab's strongest setting pushes harder.

In the modes that offer aiming, you can aim this lever at particular parts and pathways of a network. Otherwise it applies to the whole network.

The pressures it answers

No pressure page lists this lever among the levers that answer it.

Where you can pull it

The evidence behind its effects

The Lab cites these claims from the evidence registry for this lever's effects.

A 2025 survey-vignette experiment in Estonia reported that citizens' intended use of a government chatbot relates to perceived usefulness and trust in the technology, that privacy concerns relate to service-provision uses but not to information-provision uses, and that trust in government, explainability, and the amount of information provided were not related to intended use.[†]

alishani2025GroundingAcademicSave

Alishani, Homburg, When citizens meet the chatbot: Evidence from a survey vignette experiment in Estonia (Public Policy and Administration, 2025) https://journals.sagepub.com/doi/10.1177/09520767251404286

https://journals.sagepub.com/doi/10.1177/09520767251404286

Grounds: model org: burokratt_estonia

Pull this lever in the PAN Lab and watch which way it pushes the network.

Open the PAN Lab