Not by itself if it's naive, but if it's able to assess target health and avoid degraded instances then it becomes a component in HA, the other being integrating an orchestrator for gracious recovery.

Combined with a replication strategy and automated health checks, a load balancer could direct traffic to a healthy instance automatically.

What happens when the load balancer fails?