(When) Does Funding Corrupt?

Academic integrity and intellectual distortion

There’s some interesting discussion on Daily Nous about an open letter urging stricter “conflict of interest” disclosures for philosophy journals, especially for authors with ties to the AI industry. Relatedly, I previous saw philosophers on social media expressing concerns to the effect that whenever they see a colleague offer a “pro-AI” take, they wonder whether they’re on an industry payroll. There seem at least two distinct concerns floating around:

(1) External funding may be biasing individual philosophers, incentivizing them to lie for profit or at least engage in motivated reasoning that distorts their judgment.

(2) External funding may distort the literature by “flooding the zone” with work from (even reasonable and sincere) “aligned” researchers, resulting in distraction and agenda capture compared to what’s objectively warranted.

I’m a bit dubious about whether either concern really makes much sense in relation to academic philosophy—there’s a kind of epistemic transparency to our work in that everything you need to assess it is right there on the page; there’s no such thing as a fraudulent thought experiment. But I’d welcome correction if I’m missing something.

Are there financial incentives to reach certain verdicts?

The most obvious source of academic corruption is when one has financial incentives to publish certain pre-determined verdicts. I think an indirect form of this is ubiquitous: we all know you’re more likely to get a job in academia if you work on faddish topics and spout the generally-approved political and naturalist orthodoxies, for example. But I don’t see much scope for direct financial incentives in philosophy, even from external funding. At least in my experience, a typical research grant doesn’t provide any extra income;¹ it just provides supplemental work-related resources like teaching relief or conference funding to do more research on a certain topic than you otherwise could have. There’s no obvious incentive for a tenured professor to pursue that except insofar as it is work that they independently want to do. (Matters may be different for the precariously employed, desperate for extra publications on their CV. But they’re also less likely to publicly express controversial opinions, for fear of offending future hiring committees.)

It may be worth flagging here that, unlike in the sciences, there’s no expectation (at least in American universities) that philosophers pursue or secure external grant funding. Stronger still, I don’t even see much benefit to securing such funding except insofar as it helps you to pursue a project that you independently value more than teaching. As a result, I don’t really see much potential for motivating researcher “bias” here.²

Consulting jobs, by contrast, provide extra income. But they won’t always change speech incentives. For example, like many philosophers, I’ve done some side work (of the sort Dan Greco describes here) creating training data to improve AI’s philosophical reasoning. But I don’t even know which AI labs end up using the training data, and I can’t imagine that anyone involved in the project cares in the slightest about the personal views of the contributors. So I also can’t see how this sort of consulting could be at all biasing.

So I’m not actually aware of any financial incentives for philosophers to adopt “industry-friendly” views. If others know something I don’t, I’d welcome hearing it. (I don’t know what happens when academics are directly hired by AI labs—what those jobs involve or what constraints they might place on one’s extramural speech. If anyone with direct experience can shed light on this, I’d be very curious to learn more.)³

I wonder what those with the first concern are imagining—especially those panicking about the occasional “pro-AI” take on social media. Do they think AI companies are hiring academic philosophers to lie about their views, paid per comment? Or that doing any sort of consulting work will automatically dispel past concerns and turn a critic into an uncritical cheerleader for the industry? The concrete mechanism being imputed isn’t clear to me. (I’m also not aware of any academic philosophers who could reasonably be characterized as “uncritical cheerleaders” for the industry. Are you? I can point to plenty of blind dismissal that seems very clearly ideologically motivated, on the other hand!)

Do external funders increase or decrease distortion?

The second concern—about distorting the overall literature—makes (slightly) more sense mechanistically, but I think may be normatively harder to defend. Ideologically motivated research grants may marginally “boost” a specific perspective by providing aligned researchers with the time (via teaching buyouts) to publish more on the specified topic. Or they might fund postdocs for aligned researchers who might otherwise have dropped out of academia. But it isn’t clear to me why we should regard any of that extra support for philosophy as a bad thing. (In general, more philosophical research seems good to me, even though some particular instances can be misguided.)

Concern about external influences “distorting” the status quo balance of academic attention assumes that we have some reason to think well of the status quo balance. I’m not sure that we do. I don’t know that philosophers in general have especially good research taste. Which topics become “fads” has (to my eye) little discernible correlation with objective interest or importance. I expect it’s pretty rare for academics to actually reason themselves into their political positions, as opposed to “laundering vibes” or rationalizing the position that’s determined for them by social desirability bias, though of course their resulting arguments still need to be engaged with on their merits. (Just flagging that I don’t think we should defer to popular academic views as reliable by default. And the direction of error may well be systematic rather than random noise.) The standard political ideology of academics, while better than the most prominent alternative, still strikes me as misguided in many ways. Insofar as there’s an uncritical intellectual and political monoculture in academia, external influence—including industry funding—could plausibly serve as a corrective⁴ by supporting meritorious work that the academic monoculture is unwarrantedly neglecting or even suppressing.⁵

(I don’t know that there’s any way to adjudicate that question independently of one’s first-order assessments of the merits of various positions. Not every viewpoint merits scholarly representation, of course. But I do think academics should take political bias more seriously. There are some pretty striking recent examples of the monoculture “flooding the zone” to the exclusion of obviously more important issues, after all.)

Do you know anything about the politics of “Big AI”?

Something I find especially striking is how many philosophers will speak of “AI companies” and “industry interests” as a monolith, apparently unaware that the two leading AI labs have diametrically opposed politics. OpenAI’s unofficial super PAC, Leading the Future, is opposed to industry regulation and spent millions opposing Alex Bores’s congressional run. (They also apparently funded false-flag sockpuppets to discredit the AI safety movement by attributing to them unhinged, violent “supporters”.)

Anthropic, by contrast, seems genuinely committed to AI safety, has consistently called for thoughtful regulation and oversight of the industry, and unofficially spent millions supporting Bores.⁶ They also famously gave up a lucrative Pentagon contract (and suffered significant political retaliation from the Trump administration labelling them a “supply chain risk”)⁷ rather than grant the Pentagon a blank check to use their AI models for autonomous killings or domestic surveillance. (OpenAI and Google both swept in to fill the gap.)

So the common assumption that “industry funding” necessarily means “anti-regulation” reveals remarkable ignorance of the central political conflict within the AI industry.

Correcting the biggest biases

While I remain dubious of the case for COI disclosures in philosophy, I don’t mind sharing my own disclosures (as I have at relevant points in this post). I disagree with Victor Kumar that “If you find out someone’s research is funded by the Mellon Foundation, its credibility slips.” Mellon and Templeton each select (in different ways) for perspectives that systematically differ from my own, so I can predict disagreeing with the work (at least in emphasis). But if I read the paper, I’ll form my own judgment on it. I don’t think the funding information should affect my philosophical judgment in the slightest. The arguments stand or fall on their own merits; the authors could be dogs (or LLMs) for all I care. (Again, it’s not as though philosophical thought experiments are subject to p-hacking or other forms of background research fraud.)

Following the argument, or distracted by ad hominems?

If we really care about distortions threatening our collective academic integrity, my top recommendation would be to look out for areas of false consensus and social pressure to conform. Then work on being more tolerant of disagreement and curious about research that doesn’t immediately “vibe” with orthodoxy—especially while serving on hiring committees.

I worry about our more ideological colleagues pouncing on any excuse to dismiss or delegitimize disagreement. In my experience, it’s already rare for philosophers to be willing to invest in serious cross-camp engagement. The last thing we need is more excuses to avoid hearing each other out.

1. Unless some provide summer salary? My external grants to date haven’t. (I’ve been lucky to get a few internal UM grants for summer projects, though.)

2. For example, you can see from my CV that I’ve received several grants from various EA-affiliated organizations to work on projects like utilitarianism.net and my currently-in-progress textbook on the ideas behind Effective Altruism. The teaching buyouts from these grants allowed me to get a lot more written than would otherwise have been possible. Indeed, I most likely wouldn’t have undertaken these textbook projects at all without the support, since I’d rather spend my limited “personal”/discretionary research time on cutting-edge ethical theory like Beyond Right and Wrong. So I don’t know that it was in my interest to work on the externally-funded projects, except insofar as I found the work to be more meaningful and worthwhile than teaching those semesters would have been. (My wife suspects that, given widespread anti-EA and anti-utilitarian animus in the discipline, the public association is more likely bad for my career, which seems a shame. I think the resulting work has been objectively pretty good!)

3. If AI labs seem more likely to hire academics who publicly adopt specific views, that might create some incentive for people who hope to be hired by them in future to position themselves accordingly. But that’s not the sort of thing that any COI disclosure could possibly uncover. (I’m also not aware of any evidence that AI labs have been selecting for specific views.)

4. On the topic of AI specifically, Victor Kumar suggests that “philosophers are indeed flooding the zone—but with anti-AI research…”

5. As I’ve said before, “we might well expect philanthropic funders to be guided by better priorities than those of typical academics. (Imagine if Doctors Without Borders funded medical research.)” And while industries will have their own interests, I’m not aware of any a priori reason to expect those to be anti-correlated with the good. YIMBYs may make common cause with real estate developers, for example. Industry can be good.

6. Both companies officially disclaim the political activities linked to their funding.

7. Hegseth initially tried to declare that “no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic.” It soon emerged that he only had authority to bar them from doing so in the course of their Pentagon work, not more generally (and even this is undergoing legal challenge). Still, the “supply chain risk” designation was an unprecedented attack by the US government on an American company. See Ezra Klein’s interview with Dean Ball on ‘Why the Pentagon Wants to Destroy Anthropic’.