It's difficult to comment without some specifics of the programmes you're talking about, but there are a few important things to bear in mind, especially when comparing the U.K. and the US.

First, the U.K., while it has its own history of racism, didn't practice de jure segregation or have the same relationship with slavery on its shores. The UK's divisions have always been more about class than skin colour.

Following from this, the areas of greatest deprivation, mainly in the North, where majority white areas. The mining towns where people lived in immense poverty and worked in awful conditions (and then lost their jobs under Thatcher), for example, where almost entirely white. So programmes aimed at these areas would benefit white people pretty much exclusively.

But this brings us to another important difference between the UK and the US, namely demographics. While Black people are around 13% of the population in the US, they're only around 3% in the U.K. While white people are around 86% (around 70% in the US). So if you're looking purely in terms of raw numbers, a greater number of white people will benefit from pretty much any effort you make to reduce poverty. You have to look at proportions to see whether you've been effective in a "race-neutral" way.

And lastly, white woman seem to have been the greatest beneficiaries of affirmative action in the US. They've certainly made the greatest advances in terms of employment and education while affirmative action has been in existence. So even when you take race into account, if you aren't careful about your implementation, "white supremacy" still creeps in.

I think the best way to practice affirmative action would be to forget about racial representation entirely and look at districts. Which areas have the highest crime? The poorest education? The highest levels of poverty? I'm wiling to bet that these districts would overlap pretty consistently with each other. And, indeed, with redlining maps. People would disproportionately benefit from efforts made here if done well. Then you'd have some areas that were in rural America or in the Appalachians and were majority white. Great, invest resources there too.

Even gentrification doesn't need to be a curse for poor neighbourhoods. It's about whether the increased revenue goes back into the community or not. And what metrics you use to judge an area's affluence. None of this is simple, of course. But I think it's a far better way of thinking about helping the underprivileged.