Gemini AI controversy highlights AI racial bias challenge

title placeholder | GZERO AI

Marietje Schaake, International Policy Fellow, Stanford Human-Centered Artificial Intelligence, and former European Parliamentarian, co-hosts GZERO AI, our new weekly video series intended to help you keep up and make sense of the latest news on the AI revolution. In this episode, she questions whether big tech companies can be trusted to tackle racial bias in AI, especially in the wake of Google's Gemini software controversy. Importantly, should these companies be the ones designing and deciding what that representation looks like?

This was a week full of AI-related stories. Again, the one that stood out to me was Google's efforts to correct for bias and discrimination in its generative AI model and utterly failing. We saw Gemini, the name of the model, coming up with synthetically generated images of very ethnically diverse Nazis. And of all political ideologies, this white supremacist group, of course, had few, if any, people of color in them historically. And that's the same, unfortunately, as the movement continues to exist, albeit in smaller form today.

And so, lots of questions, embarrassing rollbacks by Google about their new model, and big questions, I think, about what we can expect in terms of corrections here. Because the problem of bias and discrimination has been well researched by people like Joy Buolamwini with her new book out called “Unmasking AI,” her previous research “Codes Bias,” you know, well established how models by the largest and most popular companies are still so flawed with harmful and illegal consequence.

So, it begs the question, how much grip do the engineers developing these models really have on what the outcomes can be and how could this have gone so wrong while this product has been put onto the markets? There are even those who say it is impossible to be fully representative in a in a fair way. And it is a big question whether companies should be the ones designing and deciding what that representation looks like. And indeed, with so much power over these models and so many questions about how controllable they are, we should really ask ourselves, you know, when are these products ready to go to market and what should be the consequences when people are discriminated against? Not just because there is a revelation of an embarrassing flaw in the model, but, you know, this could have real world consequences, misleading notions of history, mistreating people against protections from discrimination.

So, even if there was a lot of outcry and sometimes even sort of entertainment about how poor this model performed, I think there are bigger lessons about AI governance to be learned from the examples we saw from Google's Gemini this past week.

More from GZERO Media

The Liberian-flagged tanker Ice Energy, chartered by the US government, takes Iranian oil from Iranian-flagged Lana (formerly Pegas) as part of a civil forfeiture action off the shore of Karystos, on the Island of Evia, Greece, in May 2022.
REUTERS/Costas Baltas/File Photo

The Trump administration is reportedly considering a strategy to disrupt Iran’s oil exports by stopping and inspecting Iranian oil tankers at sea. The US would use the Proliferation Security Initiative, established in 2003 to prevent the trafficking of weapons of mass destruction, as a legal justification for the inspections.

Donald Trump issues a proclamation from the Oval Office
REUTERS/Kevin Lamarque

US presidents don’t typically talk to organizations the US government has labeled terrorist groups, but Donald Trump is not a typical US president.

President Donald Trump addresses a joint session of Congress at the US Capitol on March 4, 2025.

Win McNamee/Pool via REUTERS

You didn’t need to sit through all 99 minutes of Trump’s peroration to know that he gave himself an A++ on his first six weeks in office, writes GZERO Publisher Evan Solomon. But if Trump gets to grade himself, maybe it’s time for a more objective report card — one that looks at two criteria: Trump as a dealmaker and Trump as a manager.

The Energy Security Hub at the 2025 Munich Security Conference featured in-depth discussions on energy innovation, security, and market viability. Fatih Birol, IEA executive director, discussed growing global energy demand, especially the rapid rise in electricity outpacing overall growth. He noted electricity demand is projected to increase six times faster than total energy in 10 years, underscoring the need for electrification and grid expansion. As energy systems become decentralized and digitalized, the CEO of E.ON, Leonhard Birnbaum, said: “You’re either fully digitized – or you’re done.” Key takeaways: Energy security requires developing and securing electricity grids Technological openness is a unifying element for getting to net zero Bridge the “Valley of Death” to scale markets New global partnerships will help Europe stay competitive Public acceptance will strengthen democracy You can read the full Executive Summary from the BMW Foundation here.

a crowd of people outside of a white building

In a 5-4 split decision, the US Supreme Court on Wednesday ordered the Trump administration to disburse nearly $2 billion in foreign aid funds for work completed by contractors and grant recipients under the US Agency for International Development and the State Department. Does this tell us much about how the top court will handle future Trump-related cases?

Canada's Prime Minister Justin Trudeau waves as he leaves after testifying at the Public Order Emergency Commission in Ottawa, Ontario, Canada, on Nov. 25, 2022.

REUTERS/Patrick Doyle

Justin Trudeau’s tenure as prime minister of Canada winds down this weekend, as his Liberal Party chooses a new leader and gets set for a transition of power. His near-decade as PM began in the fall of 2015 when he embodied youthful optimism and a progressive agenda. We look at his biggest achievements and controversies.