Effective altruism (EA) is a philosophical and social movement premised on the idea that people who want to help others should use evidence and reasoning to identify which interventions produce the greatest good per dollar or hour invested, and then direct their resources accordingly. Originating in the work of philosophers Peter Singer and Toby Ord, and institutionalized through organizations like GiveWell, the Open Philanthropy Project, and the Centre for Effective Altruism, it became one of the most influential frameworks in global philanthropy by the 2010s. At the collective scale, EA has directed hundreds of millions of dollars toward interventions in global health, animal welfare, biosecurity, and artificial intelligence risk. Its strengths are genuine; its critiques are structural.
The case for EA begins with a moral arithmetic that is hard to dismiss. If a wealthy individual in a rich country can save a child from dying of malaria by donating the cost of a dinner out, and if they decline to do so without compelling reason, something morally significant has occurred. Singer's drowning child argument — that proximity should not determine moral obligation — remains philosophically robust. EA's insistence on evidence-based giving has exposed the waste and ineffectiveness of much of the charity sector, producing consumer-like pressure on NGOs to demonstrate measurable impact. Organizations like GiveWell have created transparent, publicly accessible research into intervention effectiveness that did not previously exist and that serves both individual donors and institutional philanthropy.
But at the collective level, EA's structural choices produce distortions that compound in ways individual-level analysis obscures. The first is the problem of tractability bias. EA's quantitative framework systematically favors interventions that are measurable over those that are important. Distributing malaria nets has measurable outcomes; dismantling the trade policies that keep African countries poor does not. This is not merely a methodological preference — it shapes where massive philanthropic capital flows. The collective result is a philanthropy infrastructure that funds technical fixes while systematically undervaluing political and structural change.
The second distortion is what critics have called the moral laundering problem. EA's framework evaluates giving independently of earning. A hedge fund manager who earns high returns from extractive industries and donates a significant portion to GiveWell-recommended charities is celebrated within EA for their effective generosity. But the earning and the giving are not separable at the systemic level. Capital extracted from low-wage supply chains flows through financial instruments managed by firms that the EA donor works for, producing the wealth that funds the bednets that partially compensate for the global health effects of the trade and labor policies that the firm's portfolio companies lobby to maintain. The collective accounting that EA declines to perform is precisely the accounting that Law 1 demands: genuine connection requires seeing the whole circuit.
Third, EA's prioritization methodology — particularly in its "longtermist" variant — raises distributional concerns at the collective scale. Longtermism holds that the welfare of vast numbers of future people should dwarf consideration for present suffering. This framework has directed enormous EA-affiliated capital toward AI safety research, pandemic preparedness, and existential risk reduction — concerns that, whatever their validity, are primarily salient to the professional and technical class that EA disproportionately recruits. The implicit claim is that wealthy, technically sophisticated people reasoning about the long-term future are the appropriate custodians of civilizational risk. Critics including Émile Torres and Timnit Gebru have argued that this framework is both epistemically overconfident and structurally reproduces the authority of existing elites.
Fourth, EA's relationship to power is one of deliberate disengagement. The movement has generally avoided political engagement, legislative advocacy, and movement-building in favor of technical interventions and individual behavior change. This is partly philosophical — the EA framework is skeptical of political action's tractability — and partly sociological: the movement's donors and institutional bases are uncomfortable with political confrontation. The collective consequence is that EA reinforces a model of social change in which individuals with resources make technical decisions about resource allocation, while the political and economic structures generating the problems that EA addresses are left intact.
The FTX collapse in 2022 added a new dimension to the critique. Sam Bankman-Fried was the most prominent practitioner of "earn to give" — the EA-endorsed strategy of maximizing earnings in order to maximize donations. His fraudulent enrichment of himself at the expense of customers was subsequently defended by some EA figures on utilitarian grounds before the full scope of the fraud became clear. The episode exposed how a moral framework built around outcomes can generate motivated reasoning that licenses harm in the present for hypothetical benefit in the future.
EA's genuine contributions — evidence-based evaluation, global moral concern, the disruption of donor self-congratulation — remain valuable. The task at the collective level is to integrate these contributions into a framework that also accounts for structural causes, distributional politics, and the connection between how money is made and where it goes.