The US Department of General Services (DOGE) has begun using Meta's AI model Llama 2 to process federal employee responses. Rising security concerns have prompted legislators to demand an investigation.
Use of Llama 2 in Federal Operations
Llama 2 has been employed by the Department of General Services to manage email categorization. Despite newer models Llama 3 and 4, DOGE opted for Llama 2 due to accessibility constraints. Involvement of Elon Musk's Grok was not possible as it was exclusively proprietary.
Congress Calls for Transparency
The use of this AI model has prompted inquiries from legislators. Over 40 members of Congress have called for clarification regarding DOGE's dependency on Llama 2, highlighting the need for traceability in the deployment of AI models across federal agencies.
Historical Concerns About AI in Government
Concerns about government AI practices have been noted as far back as the 2010s, with scrutiny surrounding transparency and data security in public sector AI usage.
The deployment of AI in government processes necessitates careful attention to security and transparency issues, particularly in light of ongoing demands for oversight and accountability.
