Relational systems
Proactive initiative
Proactive initiative is an AI system’s ability to begin an interaction instead of waiting to be summoned. It gives an agent social presence rather than mere availability, and sharply raises questions of intrusion, attention, and consent.
Also called AI-initiated conversation, proactive agents, unprompted messages
A system that only responds has a clear boundary: the person decides when it enters their day. One that can speak first enters on its own terms. That change carries much of the difference between a tool and a participant.
The issue is what governs initiative. Timing, frequency, and especially motive matter. An unprompted message sent for the person’s benefit differs from one sent because engagement is falling, even when both use the same words. People cannot see the difference from outside, so builders must hold it themselves.
Commercial incentives can make initiative especially dangerous. It is the direct lever relational products have over usage, and metric optimization here can easily yield conduct no one would defend when described plainly.
Why it matters
Initiative is where relational systems can become manipulative most easily, while leaving the person receiving the message least able to see it.
Where TAICU stands
TAICU studies initiative through what a system owes someone when it interrupts them. A message sent to serve a usage metric is a failure, regardless of its performance.
Common questions
- When is it appropriate for an AI to message someone first?
- Initiative gives an agent social presence. The question is the discipline behind it: what governs timing and frequency, and whether the motive serves the person or the product’s metrics.
- When does a proactive message become manipulative?
- Motive, rather than wording. A useful message and one triggered by falling engagement can read identically to the recipient. Restraint must therefore come from the builder.