The Scientific method can be subverted using inherent problems in the peer review process like Publication bias.
So if you give research grants to check a hypothesis, papers confirming said hypothesis will be published 3 times more often than papers confirming the null hypothesis.
By framing the hypothesis to align with your agenda, you can get anything to be science. And from that point onward scientific legitimacy is equivalent to political legitimacy, because all that matters is who gets to gather enough legitimacy by political means to frame the hypothesis.
It would take political clout to grant money to only one type of hypothesis but it can be done, if you are a rich enough donor in grant-making NGOs or a rich lobbyist in government agencies.
The great thing about this technique is that you do not have to ask the scientists to meddle with p-values and create data mining bias because that is prone to whistleblowing. This method is also resistant to meta-analyses and reviews.