Mutualisms have frequently been referred to as “reciprocal exploitation”. I would argue this term is unhelpful for several reasons: first, it doesn’t best capture how the vast majority of mutualisms evolve. Second, it comes with unhelpful semantic baggage, muddying the waters when it comes to understanding mutualism evolution.
Taking a very liberal definition of exploiter - an individual that benefits at the expense of its partner (see below for an alternative definition)- it is true that within a mutualism, both partners benefit at some cost to the other, but the costs involved in exploitation versus mutualism are often quite different and asymmetric between partners.
Let’s consider a bee that visits a flower for nectar. One could say that by consuming nectar, the bee is exploiting the plant. However, the production of nectar by the plant is meant to attract pollinators - it serves no other purpose. Therefore, the bee is merely consuming a resource that was intended for it. From the bee’s perspective, moving around from flower to flower is arguably costly, however, this cost is unavoidable if the bee must search to find food, let alone nectar. Is the plant really exploiting the bee’s ability to fly around, even if it to pays the bee to do so via nectar? Therefore, I would argue, rather than reciprocal exploitation being demonstrated here, it's that the plant capitalizes on the bee's propensity to move about in search of food.
Now, I am not saying exploitation does not occur within a mutualism context. Using a more strict definition of an exploiter - an individual that exploits a resource or service intended for another, we can think of several examples. The first relates to the plant-pollinator mutualism described above. Some species of bees are nectar robbers - they steal nectar from a flower, often by drilling a hole into the flower’s corolla. Because they don’t come into contact with the flower's anthers and stigmas, these bees play no role in pollination. Not only do they incur direct costs to the plants via tissue damage, nectar robbers deplete a resource that would have otherwise been consumed by a true pollinator. Some exploiters actually evolve deceptive traits that are costly to the interacting partner. For example, some species of plants produce flowers that mimic a female pollinator, both in terms of morphology and scent - when a male tries to mate with the "female", it unintentionally pollinates the flower. This form of deception could be costly to the male in terms of the time and energy spent trying to mate without the possibility of success.
So far, I’ve made a case for why semantics matter - when we call a mutualism reciprocal exploitation, we disregard how some costly resources or services are meant to be utilized (e.g., nectar) while others are almost unavoidable (e.g., foraging bouts to find food), and fail to delineate real exploitation that sometimes occurs within a mutualism. Now, I am going to point out when our semantics are less than useful. Let’s return to the bee example, except that this time, the bee actually consumes a portion of the pollen it collects, providing the pollen to its offspring as a source of food. Obviously, pollen did not evolve with the intent to feed bee babies, so should this bee be considered a mutualist, or an exploiter? Whether a pollen-collecting bee is an exploiter or mutualist depends on the net benefit it provides to the plant: if it consumes too much pollen while delivering very little, it may be considered an exploiter, whereas if the amount of pollen it successfully delivers outweighs the amount consumed, it can be considered a mutualist.
To use another example, fig wasps are technically seed exploiters; they lay their eggs within the seeds of a fig, thus providing their offspring with nutrition upon hatching. Fig wasps are costly to the fig, because had the fig not been exploited by the wasp, it would have produced more seeds. However, figs have found a way to turn the exploitative behaviour of their partner into a net benefit by ensuring every new generation of female fig wasps that emerge from the fig are plastered with pollen. As these females enter a new fig to lay their eggs, the pollen they carry gets transferred, and thus, the fig gains a reproductive benefit.
What’s really intriguing about mutualism is that even when the cost endured by one partner is not intended (e.g., pollen and seed consumption), evolution may allow such an individual to find a path forward - that is, gain a net benefit from the interaction. It may be for this reason that mutualisms are ubiquitous in nature. However, this point also calls into question why some interactions are or become mutualistic while others remain largely exploitative.
TLDR: The problem with describing a mutualism as reciprocal exploitation is that it is often the case only one partner is motivated by resource exploitation, while the other evolves ways that make this exploitation work for them. It is more helpful to think of mutualisms as interactions in which one or both partners learn to make lemonade from lemons.