Peoplestayin charge
Privacy, control and responsible use are part of how Aftaris is introduced into a product, not a policy written afterwards. These commitments apply wherever it is used.
a person decides whether to proceed.
Five commitments, and what each one means.
Aftaris does not take consequential actions on its own authority. Where it suggests one, the person using the product decides whether to proceed.
Privacy
Aftaris is designed to minimise unnecessary use of personal information and to respect the data boundaries of each product.
Control
People and product controls stay responsible for sensitive actions. Aftaris does not take them on its own authority.
Evaluation
New capabilities are evaluated against real product needs before they are deployed more widely.
Usefulness
Aftaris is developed around work people actually do, rather than demonstrations built to showcase AI.
Accountability
Important AI-assisted actions stay visible, reviewable, and governed by the product they happen in.
What that looks like day to day.
Human oversight
Aftaris is designed to assist people and improve workflows, not to remove human judgement from decisions that matter.
- Sensitive or consequential actions can require explicit human confirmation, depending on the product and workflow.
- Where Aftaris suggests an action, the person using the product decides whether to proceed.
- Material decisions about people are not left to an automated experience alone.
Privacy
Aftaris is designed to minimise unnecessary use of personal information and to respect the data boundaries of each product.
- Aftaris operates within the permissions of the product it is used in.
- Data practices depend on the product and use case, and are governed by that product’s privacy policy.
- Sensitive use cases require appropriate safeguards before an Aftaris-supported experience is introduced.
See our privacy statement for how this site and this company handle data.
Limitations
Aftaris is in active development, and we would rather be clear about that than overstate it.
- AI-assisted answers can be incomplete or wrong, and should not be treated as professional, legal, medical or financial advice.
- Where a product deals with regulated or official information, the official source stays authoritative.
- Capabilities differ between products, and not every capability is available in every experience.
- We do not claim capabilities Aftaris does not have.
Product controls
Aftaris is deployed inside products that carry their own controls, and those controls stay in force.
- Access, permissions and audit remain the responsibility of the product in which Aftaris is used.
- Important AI-assisted actions stay visible and reviewable within that product.
- Product owners decide which Aftaris-supported experiences are enabled for their users.
Responsible deployment
New capabilities are evaluated before wider deployment, and introduced progressively rather than all at once.
- Aftaris is developed and evaluated against real product needs.
- Capabilities are introduced into live products one at a time, as they are tested and considered ready.
- An experience that does not meet the standard we expect is not released.
Feedback
If you believe an AI-assisted experience in one of our products behaved incorrectly or unfairly, tell us. Email Aftermath Ltd. with the product name and a description of what happened, and we will look into it.
Aftermath Ltd. · Malta · European Union · Last updated 28 August 2026
