Hereafter devised

There’s a certain shape of clause that has been enabling the expansion of scope in creative contracts for nearly a century. But something that became boilerplate in a few industries has spread into many others, and with the way creation and innovation cross-pollinate across multiple spaces, it has the capacity to shape the scope you haven’t even started yet.

It exists in variations of “in any medium or format now known or hereafter devised” and it’s the kind of thing where scope creep can become very real for your work, and with the technologies available now, even aspects of what makes up your identity.

This style of clause has been around for nearly a century. In 1930, Hans Bartsch assigned the rights to reproduce work “visually and audibly by the art of cinematography or any other process analogous thereto.” But in the case of Bartsch v. Metro-Goldwyn-Mayer, Inc. (1967), Bartsch’s widow argued that MGM licensing the work for television — which was in the 1930s considered experimental — was not analogous to cinema. She lost that case, but the ambiguity of the word ‘analogous’ opening up the potential for dispute contributed to the broadening of legal clauses for future-protection.

There was a period where rights were tied to broadcast mediums, but with the advent of home video, Cohen v. Paramount Pictures Corp. (1988) argued that VHS/Betamax is not television, and Tele-Pac, Inc. v. Grainger made the case that the user-controlled showing of media was not a broadcast medium.

Less than a decade later, the internet and the accessibility it afforded brought about a shift in the legal language for music, publishing, and journalism leading to the expansion of the clause to encompass digital and electronic formats as well (F.B.T. Productions, LLC v. Aftermath Records (2010); Random House, Inc. v. Rosetta Books LLC (2001); New York Times Co. v. Tasini (2001)).

With every innovation the scope and definition of what is legally covered has changed, because who knows what technology, as it is now known, will become next. The clause is so well-trodden that entertainment lawyers joke about why we should limit ourselves to “throughout the universe” when we could cover parallel ones as well.

Up until now, most of that legal infrastructure captured creative work product and the ‘hereafter’ that was anticipated related to distribution. But now, the person is the product, not just their work. And the coverage is not just for distribution, but also derivatives.

Karen Jacobsen has one of the most recognised, and utilised, voices in the world. She took on a job for a one-time payment with a contractually broad scope that led to her voice being used in millions of GPS units, PA systems, and the iPhone. This was in the late 1990s-early 2000s and while I don’t know if her contract used that specific language, it certainly had the range part covered. And more than 20 years on, we are still just as exposed to legal breadth of scope.

In 2023, one of the points of dispute during the SAG-AFTRA strike was about studios taking 3D scans of background actors able to be used “throughout the universe in perpetuity in all media whether now known or hereafter devised” (Variety, 2023). What was once considered meaningless boilerplate is a lot more meaningful with technology that could mean an actor may never need to be paid for the use of their image again.

The power asymmetry between those who set the conditions and those who are subject to them is also why the precedents set by established artists are important, even if we’re not in the creative industries. Samuel L. Jackson crossing out broad clauses at contract stage (THR, 2023), Matthew McConaughey, Taylor Swift, and the Backstreet Boys trademarking their voice and image (Greenspoon Marder LLP, 2026; BBC, 2026; Billboard, 2026) are all building the foundations that the rest of us need to be able to defend the various aspects of our identity as well.

Even if you consider yourself a regular person outside the creative industries, you aren’t immune to this type of exposure.

ARN Media cloned the voice and likeness of one of their finance team members using ElevenLabs and turned her into an AI host named Thy for CADA, a Sydney radio station (AFR, 2025). Beyond ARN Media’s failure to declare the use of AI for six months — itself a potential violation of ElevenLabs’ own usage policy — presenter and voice artist Teresa Lim pointed out the importance of authenticity and truth in broadcast media, and raised concerns about whether the person Thy came from had informed consent over the scope of use of her clone, if she was compensated fairly for the use of her voice and image, and whether she has any say in what her clone is being made to say and do (LinkedIn, 2025; Sydney Morning Herald, 2025).

Exploitation isn’t just a theoretical risk, it is already present in our communities with voice cloning scams on the rise (SecurityBrief, 2026), and creators finding deepfakes of themselves endorsing products or organisations they have no affiliation with (Rolling Stone, 2026). Content creator Katie Steckly even found a vendor on Fiverr selling deepfake UGC videos using her likeness.

But the source of the problem isn’t restricted to small vendors on gig platforms. On July 8, Meta launched Muse Image, which can generate AI images of anyone with a public Instagram account without their consent, unless they opt out. Creative Artists Agency (CAA), whose client list includes some of Hollywood’s biggest names, called for Meta to take a consent-first approach stating, “No one’s name, image, likeness, voice, or creative work should be used by any third party, including AI models, without clear, documented consent” (Variety, 2026). The opt out mechanism itself is problematic and is reminiscent of OpenAI’s Sora 2 launch in 2025, which faced a similar backlash and forced to reverse to an opt-in model (NBC News, 2025).

Update 2026-07-10: Meta have responded to public backlash by removing the Muse Image feature (Deadline, 2026).

But the technology that enables legitimate cloning with informed consent is the same that enables fraudulent impersonation, and the clauses that enable broadly scoped use of aspects of our voice and image uses the same logic that enables the cloning of someone without their consent. The legal structures simply haven’t caught up with technologies that have developed faster than the frameworks meant to govern them.

It doesn’t help that many platforms and services that we may use regularly have also adopted the same types of legal language even if they’re not in the creative industries. What you agreed to in their Terms of Service gives them the same level of coverage while limiting the rights of their users. ToS;DR’s list of sites with “now or in the future” type verbiage includes food and beverage chains, market places, news outlets, homewares, and other services you may not have considered would need those kinds of future protections around your activities.

The other side of this debate is in setting up systems that allow people to have control over their digital doubles. In 2025, H&M announced that they were working with models and their agencies to create digital twins. The models own the rights to their twins and have control over how they are used. Other agencies are offering similar services and it offers models an additional revenue stream.

There are still many questions about how this will work. Compensation, for example, is unclear: will the models be paid more because they’re saving on travel and shoot logistics, or less because they don’t have to be physically present? There’s also the matter of contractual agreements that are traditionally in place between models and their agencies and how that will work with the technology in the middle.

Imperfect as it is, H&M and similar agencies setting up frameworks for transparent and ethical use of AI while keeping the models involved and in control is still important groundwork for better protections for all of us.

The distance between science-fiction and reality is getting smaller, but the future doesn’t have to look like Altered Carbon where our virtual twins can be copied, replaced, or owned by others without guardrails.

As long as we work to support systems and frameworks that keep the individual as part of the “hereafter devised” rather than as collateral to it, we can help set the expectations of the kind of future we want with this technology in it, and what that means for our personhood.

Stay curious,

— Michele

This article is part of the Original Works collection.