We must stop this.
And it is in our hands.

Yes. Any system can make mistakes, so we distinguish between two things. When an alert points to danger to a child, people respond immediately and protect the child, without waiting for any review. But a finding that someone caused harm, or any step against an employee, happens only after a proper human review. An insight is a signal to check, not a verdict, and the external research we are demanding will measure and publish the error rate.

The alert goes in real time to the person responsible in the setting, who steps in immediately, makes sure the child is safe and handles the situation. A serious alert is also sent, in parallel, to an external and independent body, such as an oversight authority or a dedicated call center, so that protection does not depend only on the people inside the setting. Alerts cannot be deleted or hidden, and the incident and how it was handled are logged. If there is a suspicion that a child was harmed, the duties that already exist in the law apply, including reporting to the competent authorities. Any determination about an employee's responsibility is made only after a proper inquiry, and not on the basis of the alert alone.

No. It is a tool in the hands of staff and supervisors, not a replacement for them. Kindergarten teachers and caregivers are the ones who care for the children, and supervisors are the ones who support the settings. That is why the declaration demands a dedicated budget in addition to existing oversight resources, not instead of them.

Only those who need to act: the responsible person in the setting receives alerts, and serious alerts also go to an external and independent body; the staff receive feedback on their work and pedagogical support; and oversight bodies receive a daycare-level picture of the situation. Parents do not see information about other children. Every access is logged, the information is deleted after a set period, and the Privacy Protection Authority provides oversight.

No. It detects situations that increase risk, such as shouting, prolonged crying with no response, a staff member alone with too many children, or overload, and alerts so that the pressure can be eased and help provided. It does not predict who will cause harm, and it does not rank people.

No, and this is one of the conditions in the declaration: a ban on ranking employees, and a ban on any decision against an employee based on an insight alone, without human review. The goal is to protect the children and strengthen the staff. When there is a suspicion that a child was harmed, the duties that exist in the law continue to apply, as they do today.

The Israel Innovation Authority supported and funded the development of Ein Tzofia's technology, as it does with Israeli technology companies. Such support is granted after a professional review of the project, so it indicates that the technology is serious and has been vetted. To be precise: the support was given for developing the technology, and it is not the Authority's endorsement of this initiative, the declaration, or the change to the law.

No. Signing is support for a policy change, not for a product. The Fourth Track defines a technological principle, not a product: it is open to any system that meets a strict regulatory standard, including prior approval by a competent government body, external accuracy testing before deployment and on an ongoing basis, information security and privacy requirements, and periodic independent audits. Signing does not commit you or your daycare to anything, and we will not use your details for marketing.