digital blackface

On AI, Erasure, And Accountability

As we enter 2025, my word for the year is "gently forward." This phrase has become my mantra—a reminder to move through this year with purpose, but without haste. It’s a grounding call to lean into the work that is unfolding, acknowledging both the challenges and the opportunities that source continues to bring me to and through. This year, I’m deeply committed to bringing the Soulwork platform to life, creating a space that centers Black people, particularly Black women and femmes, as the carriers of culture, intellect, and resilience.

The new year has only been upon us for 5 days, and already, we have seen ill omens.

But I’ll tell you, when I watched Meta launch and retract new AI Blackface profiles and noticed Fable's gross end of year summaries flutter across my corner of the internet this weekend, I was truly reminded of the importance of this work. But let me back up a little, and give you some context.

The Problem With AI Studio

In July 2024, Meta announced AI Studio, a platform enabling users to create custom AI characters for Instagram, WhatsApp, and Messenger. By December, Meta expanded this feature to allow these bots to generate and share content as “users” on their platforms, complete with bios and profile pictures.

Among the most controversial examples of these AI profiles were "Liv," marketed as a no-nonsense “truth-teller,” and “Jade,” described as a hip-hop enthusiast. On the surface, this looks like mere AI hype. Tech companies throwing AI solutions that no one asked for at problems that we don’t actually have. But the reaction from the public was swift and negative.

But when Karen Attiah of The Washington Post engaged with “Liv,” something surprising happened. With a few pointed questions, “Liv” started talking shit. The AI profile admitted that no Black employees were on the 10 person team that created it, citing instead 9 white men and an Asian woman, and went as far as to share internal documentation that pointed to intentional bias.

The documents reveal a series of lazy stereotypes. File names like “real talk” and a backstory of a party girl turned devoted wife and mom. The very conception of these profiles are based entirely off of oppressive stereotypes designed to subjugate and disempower Black women and femmes. We have the mammy, the Jezebel and the Sapphire all present and accounted for in this one profile. The test conversations make it clear: “Liv’s” design, like so many other AI systems, was steeped in bias from its inception. Attiah was able to get the AI to confess that its very existence “perpetrates harm.”

But this is nothing new. Black women have been sounding the alarm about technological exploitation for years. Moya Bailey, who coined the term misogynoir, has long shown how the intersections of race and gender produce unique forms of violence against Black women, both online and offline. Timnit Gebru coauthored a paper in July of 2020 revealing that facial recognition systems are less accurate at identifying women and people of color—highlighting how the algorithms behind these technologies fail to account for the lived realities of marginalized groups. Algorithms of Oppression has laid out how tech systems are steeped in racism and mirror societal inequalities. Black women continue to warn of the risks these systems pose, showing how data consumption shapes the very algorithms that impact us, and calling out how they lead to harm for marginalized communities.

These biases in AI mirror the historical oppression of Black women. Whether it’s the mammy archetype or the ‘diversity hire,’ Black women have long been expected to show up, hold everything together, and get nothing in return. AI is now the new frontier for the exploitation of Black women’s labor and likeness for the profit of white corporations. These digital systems may be new, but they’re doing exactly what corporate America has always done: stealing our image and words, actively eroding our safety and autonomy. AI-generated Blackness can’t challenge systems, demand equity, or say no to exploitation. Convenient, right?

AI-generated models, personas, and summaries are violent. They erase the need for real Black creators, thinkers, and laborers, replacing us with avatars who don’t challenge the status quo. This is a continuation of the corporate mammy: serve, smile, and stay silent.

But we are not silent. Our intellectual labor, cultural contributions, and community care have shaped every facet of this nation.

Soulwork reminds us that Black women’s labor, both physical and intellectual, has always been revolutionary. Knowing, nurturing, and expressing our souls is an act of defiance against systems that seek to commodify us. The question is: how will we hold these corporations accountable?

It was because of the outcry from Black women that Meta removed its offensive profiles, and Fable — a reading app facing similar backlash, was pressured to remove AI from its summary tools. Our voices and our buying power are powerful, but so is our peace. These incidents reinforce the urgent need for spaces that center Black women, not as products to be consumed, but as full, complex individuals whose contributions shape the world.

It’s time to refuse the role they’ve written for us, in code or otherwise. We’ve always disrupted systems that tried to erase us, and this is no different. These corporations can’t sell our culture without us. We must call them out. We must demand accountability.

And as for me, I’m pouring my energy into creating a space that centers us—not a sanitized, digitized version of who others think we are. Soulwork is more necessary than ever.

---

FURTHER READING