Membranes for FO and RO membrane should have high water permeability, high salt rejection and good mechanical resistance. Contrary to Reverse osmosis, Forward osmosis is low pressure driven process. The active layer is usually the same for both FO and RO membrane. Nevertheless, RO membrane need more important mechanical resistance in the support layer in order to resist to the high pressure. Furthermore, FO is much more sensitive to internal and external concentration polarization, in this way, there is some difference in order to limit this problem.

The membranes are constructed differently, but before you get into that you should consider the energy difference between the two.

To desalinate with reverse osmosis you need to do work on the system, because you're essentially LOWERING the total entropy. In other words, you're taking something with a relatively high entropy (i.e. 35,000 some parts per million of salt in water) and you're turning it into something with a lower entropy (i.e. 100 parts per million of salt in water plus some brine at maybe 100 parts per million of salt in water). The laws of Thermodynamics require that we ADD energy to a system to make the system more ordered (i.e. lower the entropy.) So to do this, we run pumps or electric gradients which essentially push the salty water through the reverse osmotic membrane.

But forward osmosis is the opposite. In that case, we're RAISING the total entropy of the system. We start with 35,000 parts per million salt in water and then make the system even more disordered (i.e. raise the entropy) by allowing the water in that salt solution to pass through a Forward Osmotic membrane and run in the even more disorganized playground of a 60,000 parts per million solution, for example. (This higher concentration draw solution doesn't have to be 60,000 ppm, it just has to be higher than the original solution.) But the trick here is that the 60,000 ppm draw solution doesn't have to be salt, it can be sugar or ammonia or liquid fertilizer, or anything that is less harmful than salt, or which can be removed from the higher concentrate. (For example, I can happily drink 60,000 ppm sugar solution, but not 35,000 ppm salt solution, or 60,000 ppm ammonia can be boiled out of the water and reused in the cycle.

But the key here is that since the solution is allowed to become even more disorderd than it was originally, we don't have to add energy to the system or do work on the system. It's essentially rolling downhill on the entropy gradient, and the process can GIVE us low grade heat or work, some of which we can add to the system. The reason why forward osmosis isn't commonly used yet is because it can be tough to find the right (i.e. cheap) draw solution. However it's Jim Dandy for emergency pouch desalination units on lifeboats. The user is happy to drink the high-concentrate sugar solution. These are available on Amazon for less than $30.

As for the membranes themselves, that gets technical enough that it needs formula more than words, but the construction of the membranes are similar in that they essentially act as gatekeepers to allow the osmosis to happen one direction, either against or with the entropy gradient.

The membrane will be not much different, have the same basic properties i.e. high salt rejection. Several report have been published about FO process,

Zhengzhong Zhou, Jim Yang Lee. 2016. Evaluating the viability of double-skin thin film composite membranes in forward osmosis processes. Journal of Membrane Science, 502, 65-75

Dany Roy, Mohamed Rahni, Pascale Pierre, Viviane Yargeau. 2016. Forward osmosis for the concentration and reuse of process saline wastewater. Chemical Engineering Journal 287, 277-284