Deepfakes: Can Business Uses Ever Outweigh Ethical Problems?

AI systems capable of producing synthetic voices, imagery and video content have become integrated into business operations spanning customer service through to marketing campaigns – yet the technology is simultaneously generating unresolved concerns about authenticity and permission.

The entertainment sector is proving to be a battleground for these discussions, with performers and their relatives resisting the deployment of AI to reproduce human performances without authorisation.

These systems function by training machine learning algorithms on datasets containing genuine human performances, subsequently using those patterns to create new content that replicates particular individuals.

What began as scholarly research into neural networks has rapidly evolved into commercial products that offer financial savings across numerous industries.

But these AI-generated videos have sparked vociferous opposition from employees, policymakers and relatives who find themselves face-to-face with synthetic reproductions that they never sanctioned or consented to.

Now, Zelda Williams, the daughter of performer Robin Williams who passed away in 2014, has revived appeals for individuals to cease creating and distributing AI-generated videos of her father, characterising the practice as upsetting and disputing assertions that such content represents progress.

The rise of AI actors

The discussion has grown more intense following the introduction of Tilly Norwood, promoted as an "AI actor" by its developer Eline Van der Velden.

The Dutch performer and comedian informed media she hoped for Tilly to become the "next Scarlett Johansson", a remark that triggered swift criticism from SAG-AFTRA, the US union representing actors and performers.

"It's not an actor, it's a character generated by a computer program that was trained on the work of countless professional performers," the union said in a statement.

"It has no life experience to draw from, no emotion and from what we've seen, audiences aren't interested in watching computer-generated content untethered from the human experience."

That training data concern continues to emerge across various contexts.

Most commercial deepfake systems learn by examining copyrighted material, generating questions that current intellectual property frameworks find difficult to resolve.

OpenAI and Stability AI both face lawsuits from creators who argue their work was used without permission or payment.

Actor Emily Blunt says: "That is really, really scary" on a podcast with Variety, the US entertainment publication.

"Come on, agencies, don't do that. Please stop. Please stop taking away our human connection."

Eline defends her creation in different terms: "To those who have expressed anger over the creation of my AI character, Tilly Norwood, she is not a replacement for a human being, but a creative work – a piece of art," she says.

"Like many forms of art before her, she sparks conversation – and that in itself shows the power of creativity."

Do we need deepfake regulations?

Legal frameworks have found it challenging to match the pace of commercial deployments of AI models.

The EU's AI Act, which took effect in 2024, mandates disclosure of synthetic media in particular contexts, but regulators are still determining implementation details.

Some US states have enacted personality rights legislation that continues beyond death, yet enforcement remains inconsistent and does not cover all commercial uses.

Meanwhile, China mandates watermarking of AI-generated content, though compliance mechanisms continue to develop.

In other words, no global accord currently exists.

"It's freely available to someone with very little technical skill to copy a voice, image or even a video," Arup's Chief Information Officer (CIO), Rob Greig, told the World Economic Forum.

This regulatory gap creates uncertainty for businesses weighing whether to deploy synthetic content systems.

For example, customer service operations must decide whether to disclose that callers are speaking with AI-generated voices rather than humans.

So far, social media companies have introduced labelling for AI-generated content in some cases, but implementation varies widely and detection systems often lag behind generation capabilities.

Business platforms including LinkedIn have yet to publish policies on synthetic professional profiles or AI-generated business content.

The question of consent becomes particularly complex when synthetic content involves deceased individuals. Social media now hosts tools that promise to "bring your loved ones back to life" through animated images, creating a market for AI recreations of people who cannot grant permission.

As a result, families find themselves confronting synthetic versions of relatives they never approved.

The deeply personal consequences

Zelda Williams, a film director, has experienced this ethical dilemma directly.

In 2023, Zelda supported a SAG-AFTRA campaign against AI, describing attempts to recreate her father's voice as "personally disturbing".

A recent Instagram post reads: "Please, just stop sending me AI videos of Dad. Stop believing I wanna see it or that I'll understand, I don't and I won't.

"To watch the legacies of real people be condensed down to 'this vaguely looks and sounds like them so that's enough', just so other people can churn out horrible TikTok slop puppeteering them is maddening," she writes.

"You're not making art, you're making disgusting, over-processed hotdogs out of the lives of human beings, out of the history of art and music – and then shoving them down someone else's throat."

Her criticism extends to how the technology industry frames Gen AI: "Stop calling it 'the future,'" she writes.

"AI is just badly recycling and regurgitating the past to be re-consumed. It's a waste of time and energy – and believe me, it's NOT what he'd want."