From left: Moderator Tomm Polos with panelists Liz Randall, Dor Gvirtsman, Michael Chiklis and Scott Likens at Creativity in Control, the first public event presented by the USC Institute for Actor-Driven Innovation. Photo by Katie Youn
“My voice is mine. My image is mine.”
For actor Michael Chiklis, that was the starting point. The issue was not whether artificial intelligence is exciting or inevitable, but whether the people whose work, voices and likenesses give the technology value retain control over how they are used.
That principle framed Creativity in Control, the first public event presented by the USC Institute for Actor-Driven Innovation at the USC School of Dramatic Arts (SDA). Supported by founding sponsor Adobe, the conversation brought together artists, industry leaders, technologists, students, faculty and alumni to examine what it would mean for actors to help direct technological change rather than simply react to it.
The panel featured Chiklis, the Emmy- and Golden Globe-winning actor and longtime member of the SDA’s Board of Councilors; SDA alumnus, actor and voiceover actor Dor Gvirtsman; Liz Randall, head of business operations and strategic development at Creative Artists Agency; and Scott Likens, chief AI engineer at PwC. Tomm Polos, founding director of the Institute and director of Creator Arts at SDA, moderated.
Each panelist brought a distinct perspective. Randall addressed talent rights, contracts and the protection of performers’ digital assets from within the entertainment industry. Likens approached the conversation as a technologist who works with AI across industries. Chiklis and Gvirtsman grounded the discussion in the experiences of working actors at different stages of their careers.
Their discussion did not begin from the premise that actors should embrace AI. It began with the reality that the technology is already affecting their work, often without their participation, and with the concern that performers need greater knowledge and leverage to protect themselves.
Ownership before opportunity
Speaking from her work protecting performers’ interests at Creative Artists Agency, Randall described control in two words: ownership and choice. “You cannot control what you do not own,” she said. Actors, she argued, should own their data and decide if, when and how they engage with technology. Consent, compensation and security must be established before any discussion of what a digital replica might make possible.
That principle informs the CAA Vault, an in-house service that captures and securely stores high-resolution data related to a client’s likeness and voice. Randall described CAA as a bank for those assets: the performer owns the dataset and can decide whether it is used or deleted. The aim is to prevent a technology company from controlling an actor’s digital identity simply because it created the scan or trained the model.
For Chiklis, the issue was immediately practical. He urged performers to pay close attention to contract language granting rights to a voice or likeness “in perpetuity” and “throughout the universe.” His own experience being scanned for a major film underscored how easily performers can lose sight of where detailed records of their faces, bodies and voices are stored and what agreements govern them.
Gvirtsman connected that need for fluency to the experience of emerging artists. Actor training remains grounded in scene study, voice and movement, he said, but entering the profession also requires an understanding of the business forces shaping it. Learning how AI functions, Gvirtsman argued, does not require endorsing every use of it. Instead, he believes that knowledge gives actors a better chance to understand the decisions being made around them and participate in shaping those decisions.
Making room for skepticism
The conversation also made room for the real fears that have made AI such a charged subject across the creative industries. Chiklis spoke directly about job and wealth destruction, particularly for artists whose entry-level opportunities may shrink as companies automate parts of the production process. If fewer people are hired for the early jobs through which artists learn their craft, he asked, how will the next generation gain the experience needed to lead the work later?
A recent graduate in the audience made the stakes personal. In the three years since graduating, they said, they had seen a production on which they worked recreated “almost word for word” with AI and posted to social media. They had also used ChatGPT to complete a complicated production form that might otherwise have left a client’s equipment stuck in customs. Their experience captured the contradiction at the center of the event: the same technology can appropriate creative labor and solve a practical problem.
Another audience member raised a different concern. While AI may enable one person to accomplish more, they worried that it could also create isolation and weaken the collaborative process at the heart of the dramatic arts. Their question shifted the discussion from what the tools can produce to what might be lost as individuals become able to do more on their own.
The panelists did not resolve those tensions, nor did they suggest that technical fluency alone could solve them. Instead, the exchange returned repeatedly to the need for artists to identify where the technology may be useful, where it falls short and where it should not be permitted.
Approaching the issue from the technology side, Likens argued that the speed of change is part of what makes the present moment so destabilizing. Technological shifts have displaced and transformed work before, he said, but historically, that has happened over decades. AI is compressing that process, leaving people and institutions far less time to absorb what is happening or adapt to it.
Likens also challenged a familiar phrase in technology circles: keeping a “human in the loop.” “I don’t want to be in the loop working for the machine,” he said. “I am the loop.” Human ideas and intentions, he argued, must remain the starting point, with technology serving as a multiplying force rather than directing the work. On a production, that could mean using AI to reduce administrative work or inefficiencies so that people have more time for creative decisions and collaboration.
That possibility did not erase Chiklis’ questions about which tasks are considered expendable, whose jobs provide essential training and whether promised efficiencies will actually be reinvested in artists. The difference between their perspectives underscores one of the Institute’s purposes: to examine both the creative possibilities of emerging technologies and the economic and human consequences of how they are developed and deployed.
Experimentation on actors’ terms
For Gvirtsman, education can offer a lower-stakes environment in which actors test those boundaries. Students rarely have the opportunity to practice performing opposite a creature that will be added in post-production, to see how motion capture changes a performance or to experiment with visual effects. Emerging tools could make rough versions of those experiences available in a classroom, allowing actors to try, fail, assess the result and build the vocabulary to advocate for themselves on a professional set.
That kind of experimentation, he emphasized, depends on the training and creative vision that precede it. AI may shorten the distance between an idea and a first draft, shot list or visual concept, but speed is not the same as artistry. “If you want to create excellent art and something that is original and is meant to stand on its own and speak to the human condition, that’s its own craft and practice,” Gvirtsman said.
Chiklis offered a similar view. He could imagine the technology helping artists realize projects that previously would have been rejected because of cost. At the same time, he did not want those savings to come at the expense of editors, performers or other collaborators whose judgment gives the work its meaning. He described himself as “of two minds,” interested in what the tools could enable and unwilling to ignore whom they could harm.
Keeping the human relationship in view
The panel repeatedly returned to the enduring value and growing popularity of live experiences, as well as the importance of human judgment and community. Likens argued that AI can mimic or simulate creativity but remains dependent on patterns drawn from what humans have already created. Chiklis suggested that the desire for shared, in-person experiences may make live performance even more valuable. Gvirtsman described the lasting relationships built when artists make imperfect work together and become comfortable failing together.
Those ideas offered a broader definition of actor-driven innovation. It is not simply the use of new tools by actors. It is an approach that begins with the capacities actors develop through their craft: presence, listening, adaptability, imagination and connection. From that starting point, the questions change. Instead of asking only what the technology can do, artists can ask who controls it, whose labor it affects, what relationships it strengthens or weakens, and whether it serves the story and the people making it.