- A fraction of solar active regions are observed to have current helicity of a sign that contradicts the polarity law for magnetic helicity; this law corresponds to the well-known polarity law for sunspots. A significant excess of active regions with the ``wrong" sign of helicity is seen to occur just at the beginning of the cycle. We compare these observations with predictions from a dynamo model based on principles of helicity conservation, discussed by Zhang et al. (2006). This model seems capable of explaining only a fraction of the regions with the wrong sign of the helicity. We attribute the remaining excess to additional current helicity production from the twisting of rising magnetic flux tubes, as suggested by Choudhuri et al. (2004a). We estimate the relative contributions of this effect and that connected with the model based on magnetic helicity conservation.