AI ethics researchers are abandoning corporate social responsibility critiques for systemic challenges to how AI development concentrates power globally. The AI Now Institute's Reframing Impact series features Timnit Gebru and Abeba Birhane questioning industry's 'AI for Good' messaging.
"It's a way to paint a positive image of AI technologies, especially in light of the backlash from grassroots resistance movements," Birhane said. "'AI for Good' allows companies to say 'Look, we're doing something good! You can't criticize us.'"
Gebru argues the dominant paradigm involves "stealing data, killing the environment, exploiting labor" as companies pursue building a "machine god." Investors pressure smaller language AI organizations to shut down when Big Tech announces competing models. "When OpenAI or Meta comes with an announcement of a big model, potential investors literally told them to close up shop," Gebru reported.
The resource concentration behind large language models disadvantages developers outside wealthy tech hubs. Smaller multilingual AI projects face funding cuts when Silicon Valley giants enter their markets, limiting linguistic diversity in AI development.
Birhane warns AI deployment may bring "surface-level improvements but also underlying destruction and division," predicting systems will "encode existing norms and stereotypes in a way that makes the rich richer and more powerful."
African governments adopting AI development rhetoric face particular scrutiny. "It's sometimes really scary the way you see some African governments jumping on the AI bandwagon and buying into this rhetoric that AI is going to 'leapfrog' the continent into prosperity," Birhane said. She notes "very little thought" goes to impacts on freedom of movement, speech, and knowledge ecosystems.
The critique challenges assumptions that AI progress benefits society universally, demanding accountability for environmental, social, and labor costs concentrated in the Global South while profits flow to Northern tech firms.