What does “risk” look like here? It means funding the research that will fail or make you look bad if you fund it.
This is true for both government and philanthropic grant giving organizations. In both cases, because People always spend money for a purpose grants are expected to have a result. In this light, derisking a grant means giving it to people and projects that the grantor thinks is most likely to produce a result.
In the case of philanthropy there is Asymmetric career risk for the employees giving the grants. If they give out grants that have no result, it will impact their career. Additionally, if it is a charity-funded philanthropy they need to raise money every year. The charity needs to show results in order to raise money. Therefore, they are incentivized to give grants that are most likely to show results.
In the case of the government, the grantors worry about wasting taxpayer money and there are various demands to show that the money is going to good use.
The least risky research is done by people and ideas with track records. This makes it hard for new people and ideas to get traction. To make new things you need to take risks. In order to get grants, people without track records are forced to work on ideas that do have track records. Therefore, the only people who can get funded to work on crazy new ideas are people later in their careers. Working in a world tends to shape your worldview around that world so people later in their careers are less likely to come up with breakthrough new ideas. (See The Structure of Scientific Revolutions.)
Another way to derisk is to give the grant to people you trust. Because of the Mere-Exposure Effect grantors will trust people who have gone out of their way to interact with the grantors or have just been in the game longer and ideas that they are familiar with . The trust aspect further serves to direct grants to people later in their careers and well-trodden ideas. Is this related to The best due diligence is network based?
The derisking incentives around grants is one of the reasons why The idea that academia is for crazy exploration and companies are for low risk exploration is bullshit and Lots of innovation is stifled because of the overhead of going off of the beaten path rather than direct forbiddance.
National Science Foundation - NSF - Grant panelists give out 6 grants for 25 applications. The people who get it are the established groups - they know the program officer, what they’re looking for.