Argentina’s PAMI adopted rules for artificial intelligence across its departments and suppliers. Projects face review before entering service, while tools already operating face continued monitoring. Decisions affecting treatment, medication or services must remain under human oversight. The framework also restricts staff to authorized platforms and sets conditions for using members’ sensitive information.
PAMI, Argentina’s institution providing social and health services for retirees and pensioners, adopted an artificial intelligence manual on November 11.[1], [2]
It applies across the institution and to suppliers handling its information. Every project bringing AI into institutional processes must receive formal review before operation. Employees using individual assistants must also choose platforms that the institution has authorized.[1]
A new AI Ethics Committee brings technical, legal, medical and information-security specialists into project review. Working with PAMI departments, it can approve an initiative, return it for changes or reject it. Oversight continues after introduction: operating tools face periodic checks and can be suspended when errors, bias or security problems appear. The committee’s work is governed by separate operating rules.[1]
Decisions affecting access to treatment, medicines or services cannot be entrusted exclusively to AI. Employees may not enter sensitive information into services lacking confidentiality guarantees. Members’ personal information should be anonymized whenever possible, with data-protection and legal approval for its use. PAMI drew on guidance from SIGEN, Argentina’s national public-sector oversight body, which keeps responsibility for decisions with people.[1]
