New applications of connectivity between vehicles and the infrastructure are developed. One of these applications is providing upstream information on variable speed limits to connected and automated vehicles. It is expected that the connectivity can contribute to safer roads due to better compliance to the posted speed limits. Those expectations are based on microscopic models with the assumption that human drivers behave similarly in mixed traffic as they do in only human driven vehicles traffic. However, few studies have shown that human drivers tend to change their driving behavior when interacting with automated vehicles in mixed traffic. For this reason, a driving simulator experiment is executed to investigate the effect of the penetration rate of connected and automated vehicles and the distance at which the information is provided on the driving behavior of human drivers. The driving behavior was analyzed in terms of longitudinal and lateral behavior in the context of a three-lane motorway. The penetration rate was found to only impact the speed adaptation when combined with a large distance of upstream information. Lower means speeds, lower section entry speeds and increased speed compliance was observed with an increasing level of penetration rate. For the effect of distance of upstream information, a similar effect was observed. When the distance was increased the mean speed lowered, section entry speed lowered, and speed compliance increased. No change was observed regarding the lateral behavior or Time Headway. As a result, it can be concluded that the cooperation between connected and automated vehicles and variable speed limits on motorways can be used to slow down unconnected vehicles more upstream, without inducing aggressive driving behavior in terms Time Headway and lane changing behavior.