Arina Pismenny & Berit Brogaard have a paper, Vices of Friendship (PDF), in which they argue against what they call a "Neo-Aristotelian" account of "true friendship". I think that what they call the 'Neo-Aristotelian' account is actually a mish-mash of different accounts that don't fit together very well, but it's worth noting why a genuinely Aristotelian account of friendship would not have the problem that they identify.
One important issue from the beginning is this notion of "true friendship". This is an ambiguous phrase. It could mean, "real friendship" as contrasted with "false friendship". It sometimes seems that Pismenny and Brogaard assume this. But it can also mean, and usually does in colloquial contexts, "friendship of the highest sort" or "friendship which most completely involves what is appropriate to friendship". Pismenny and Brogaard identify one kind of 'true friendship'. But in the former sense, Aristotle thinks there are three kinds of friendship:
(1) friendship of pleasure
(2) friendship of use
(3) friendship of excellence (also known as friendship of virtue or, as Pismenny and Brogaard usually call it, friendship of character)
All of these are really and truly friendships, in which two people are in mutually friendly relationship to each other. All of them are important for Aristotle. In Aristotle's own account, societies are primarily constituted by these friendships. The reason for this is that the best of these (what counts as 'true friendship' in the second sense) is friendship of excellence, but this kind of friendship is necessarily rare. Thus most of the friendships that constitute society are friendships of pleasure and friendships of use. We can't do without them, and an account of friendship that ignores them completely is necessarily defective. Pismenny and Brogaard do in passing acknowledge that Aristotle recognizes these by takes their labels to be pejorative. This doesn't make any sense at all. Not only are most friendships either friendships of pleasure or friendships of use, Aristotle thinks we naturally tend to develop both. (He does hold that young people tend to favor friendships of pleasure, because they overemphasize the importance of pleasure and old people tend to favor friendships of use because they overemphasize the importance of security. But this is a matter of different balances; it is not a claim that having friendships of pleasure or friendships of use is any kind of failure.) In an Aristotelian account, it doesn't make sense to read these as pejorative labels. This is going to be important for assessing the argument put forward by Pismenny and Brogaard.
In the 'Neo-Aristotelian' account Brogaard and Pismenny are considering, friendships of character have the following features:
(a) They involve mutual admiration and respect for each other's character, i.e., excellences.
(b) They involve disinterested mutual love, involving love of the other for their own sake.
(c) They involve shared life, each taking the other's joys and sorrows as their own.
(d) They involve mutual encouragement to virtue.
(e) They are each able to be role models for each other in virtuous life.
(f) They involve mutual moral criticism, each holding a mirror to the other by which the other is able to better understand themselves.
Pismenny and Brogaard argue that (b) and (d) -- and perhaps, with the former, (c) as well, and with the latter, (e) and (f) as well -- are defects in this account. The Neo-Aristotelian friendship is too disinterested and too moralistic to be true friendship.
First, they argue that, realistically, friendship love involves a desire for intimacy and reciprocity. This means that it cannot be disinterested and selfless.
This is not a problem in a genuinely Aristotelian account, because the disinterestedness that one can attribute to friendship of excellence does not exclude desires like the desire for intimacy and reciprocity. As is well known, Aristotle doesn't think you should squash desires, but that you should instead find a proper balance in how they inform your decisions. In a friendship of excellence, the friendship wouldn't exclude desire for intimacy and reciprocity; it would involve mutual moderation, finding the right proportion of it in choices to have a flourishing friendship. The sense in which one can take friendship of excellence to be disinterested is precisely the sense in which it is not, as such, a friendship of pleasure or friendship of use. The friend is not loved merely for the enjoyment they bring, nor merely for their usefulness. This does not, of course, mean that people united by friendship of excellence do not enjoy each other's company or find each other useful; rather the fundamental parameters of the friendship itself are set by good character (nobility, honestum) and not by values like the pleasant and the useful.
Given this, the second objection is perhaps more significant. Pismenny and Brogaard argue that promoting virtuous character is not necessarily a feature of 'true friendship'. The first reason is that "it is clearly unrealistic to expect that true friendships must begin with mutual admiration of each other’s virtuous character"; friendship could arise in other ways. Second, the attempt of one friend to improve the other might be resented. Third, since no one is perfect, friends are nearly as likely to lead each other in the wrong direction as not.
Again, on a genuinely Aristotelian account of friendship, none of these are serious problems. Aristotle doesn't think that friendships of excellence are the only genuine friendships; he would think it blatantly obvious that most of our friendships arise on considerations other than virtue. It is even possible that a friendship of excellence might begin as a friendship of pleasure or friendship of use. It's just that it can't be a friendship of excellence if you don't recognize each other's excellences, and recognize them as excellences. Further, while it is entirely true that the attempt to improve someone else might be resented, a fundamental element of excellence in Aristotle is finding a proper balance, one that can be determined by the virtue of prudence. People united by friendship of excellence won't be constantly nagging each other about morality; they will instead do things like offer advice, exhortation, and caution, share their own experience, and attempt to be honest, compassionate, friendly, merciful (and in other ways be virtuous) with each other. Pismenny and Brogaard seem to have the notion that mutual improvement involves beating each other constantly with morality sticks; this is simply untrue. The most serious is the third one, and I think it is a case in which Aristotle would flatly deny the argument. While no one is perfect, and being virtuous is certainly not the same as being perfect in Aristotle's account, and even virtuous people can accidentally lead each other astray, this is all necessarily accidental with respect to virtue itself. The natural tendency of virtue is to model virtue and to encourage virtue, and therefore it will do both for the most part. Friendship of virtue will no more make you perfect than virtue will (although here, as often the case, two heads will often be better than one); but it's absurd to say that it would not tend to encourage and support virtue. By its very nature, it is part of virtue's ideal habitat.
Instead of the 'Neo-Aristotelian' account, Pismenny and Brogaard give an account of 'true friendship' that emphasizes the following characteristics.
(a) It involves mutual desire to promote each other's interests insofar as they spring from core values. (Pismenny and Brogaard call this 'closeness', as in 'close friends', but they don't actually have a good argument that this is what closeness of friendship actually involves.)
(b) They involve mutual self-disclosure of emotionally intimate information.
(c) They involve being "open to a friend's direction and interpretation" (p. 245), which they call trust.
(d) They involve partially shared identity, without loss of individual identities.
They then argue that such friendships, while they may be virtuous, can also be vicious.
The obvious problem with this account is that, if we are talking about 'true friendship' in the first sense, this is obviously far too narrow. We can have good friendships that are missing any and all of these. Yes, it's probably the case that all friendships involve some kind of mutuality, trust, and sharing, but very often these will be in forms that are not specified by the descriptions given by Pismenny and Brogaard. But, on the other side, if we are talking about 'true friendship' in the second sense, it is simply wrong that these are the fully constituent features of friendship at its truest; nothing about this description would characterize either person in the friendship as most truly a friend. In fact, there is something obvious that is missing here that absolutely guarantees that meeting these criteria would not suffice to make you most truly a friend. There is no mention of friendliness. One of Aristotle's major virtues is the virtue of friendliness, and obviously the person most truly a friend will have the excellence of friendliness, to a high degree of excellence. Aristotle's friendship of excellence necessarily involves this virtue; friends by virtue are friendly to each other, where friendliness is not merely a temperamental feeling but a cultivated excellence. The (genuine) Aristotelian has a diagnosis for the problem: Pismenny and Brogaard have given us a description that, taken simply as it stands, is a description of one kind of friendship of pleasure. It is indeed truly a friendship (although far from being the only kind); it is pleasant from the perspective of our desire for intimacy and reciprocity, as well as from the perspective of our desire for shared life and our related desire to be with those who share our values, and insofar as we get involved in a friendship like this because it satisfies those desires, it is a friendship of pleasure. And they are right that it is consistent with vice, indeed quite extensive vice. Where they go wrong, the Aristotelian would say, is in thinking that this friendship that is truly a friendship is friendship in its truest form. The truest form of friendship involves friends being excellent to and with and for each other.
