Legal development

When 'Made by Humans' Becomes a Selling Point: Branding in an AI age 

Abstract image

    What you need to know

    • Faced with a deluge of AI generated works, consumers are increasingly valuing authenticity, originality and human creation, and designating products or content as 'human-made' or 'AI-free' is becoming a point of distinction.
    • Establishing a singular definition for 'AI-free' is difficult. While some contend a label should only be used on products that have been developed entirely free from AI input, because AI is integrated into so many third-party platforms, there are practical challenges to creating works without any AI involvement whatsoever. Drawing a line between human-made and AI created is therefore not straightforward and risks devaluing human created works where the AI involvement has been insignificant or limited to preparatory works or formatting only.
    • Developing a cohesive and trustworthy verification mechanism for human made products is challenging. There are many AI-free stamps and labels available for use with varying degrees of credibility, some labels are unverified self-applied badges while others may only be applied following an audited certification process.

    Books, movies, music, art and fashion. These are a few of the many industries which have been significantly impacted by AI.

    Until recently, promoting a product by its human-made or created component would have seemed absurd or illogical. Now, businesses are increasingly seeking out ways in which to promote products as 'human-made' in a market where such products and content are often indistinguishable from their AI competition.

    It is important to establish a clear and standardised mechanism to adequately and reliably communicate a guarantee of human creation. There is certainly the potential for such a scheme to exist, in the way that many stamps and labels are used on food products and household appliances (ie, vegan, organic, Fair Trade and energy efficiency ratings), but the proliferation of competing AI-free labels with vastly different standards risks undermining the credibility of the label altogether.

    Unsettled meaning of 'AI-free'

    Developing and agreeing upon a single universal definition for 'AI-free' is not straightforward. Essentially, there is a spectrum on which human made, conceived and developed products sit at one end, while wholly AI-generated (without human involvement) products sit on the other. Within this spectrum exists most created products and content, and therein lies the point of contention. Some would argue that an 'AI-free' label should only be used on products that are entirely free from AI involvement while others claim that the ubiquity of AI-integrated tools mean that taking such a strict approach would undermine or mischaracterise works where AI has played a minor or tangential role e.g. confined to tasks like formatting or noise reduction.

    Competing standards and lack of consensus

    There is currently no government-backed legal definition for AI-free but a number of different organisations are competing to create their own AI-free label. Currently, there seems to a mix of self-certification, third party verification and industry certification standards emerging:

    • Self-certification: Some businesses, authors or artists are taking it upon themselves to create stamps or labels signalling that the product or content is written, published or created by humans with no or minimal AI, such as the 'HUMAN WRITTEN' stamp applied to copies of Sarah Hall’s novel Helm.
    • Third party verification: Others are turning to third party verifications such as Not by AI and AI-free.io Human Made, or one of the growing number of verification services permitting users to download and use AI-free of labels or stamps. Auditing and assurances measures associated with these third-party verification schemes are varied and the availability of free, immediate audit-free stamps risks devaluing the use of AI-free labels.
    • Industry certification: Industry certifications are gaining traction in areas particularly impacted by AI-generated content. In the UK, the Society of Authors has developed a scheme permitting the use of a 'Human Authored' logo on books where the text is created solely by the author/authors, and the use of AI or generative AI is limited to word processing programmes, spelling and grammar checks, research, brainstorming or other uses that do not generate text. The books are then added to a register of human written works. Similarly, UK start-up Books By People has launched an 'Organic Literature' certification allowing approved books to use a fingerprint stamp. Another company called The Human Made Mark has created a 'Human Made' trust mark certifying that a film that has "humans in front of the camera and behind it". Notably, The Human Made Mark has sought to protect the 'Human Made' trust mark in Australia (although the trade mark application remains pending as at the date of this publication).

    A registered certification trade mark may be another avenue to explore. Certification marks such as the Woolmark or Australian Made logo may be used by any consumer or business to indicate that a product or service meets specific certification standards. These schemes are typically administered by an independent regulatory board or industry association with rules that have been assessed and endorsed by consumer protection regulators like the Australian Competition and Consumer Commission (ACCC).

    Australian context

    Australian businesses are similarly grappling with strategies for recognising AI-free products or content. The Australian Society of Authors has confirmed it is in talks with its US and UK counterparts about the 'Human Authored' label, but has not committed to joining forces or adopting a similar scheme in Australia at this time.

    Australian company, Proudly Human, offers a rigorous certification system to "prove the authenticity of human-created work across all creative industries", including authors and publishers, visual artists, musicians and composers, content creators, platforms businesses and brands. (see: https://www.proudlyhuman.org/). Ironically, it uses AI detection tools as well as its own methods to establish human creation of the work. Proudly Human sets a criteria for acceptable AI use for each creative industry allowing very limited AI involvement. 23 First has applied for registration of the Proudly Human stamp as a certification mark in Australia (the application remains pending), as well as Brazil, Canada, Europe, India, Japan and Korea. Joint AU Venture is also seeking certification trade mark protection for its Responsible AI logo in Australia.

    While the growing number of labels and stamps reflects a desire to preserve and protect human creativity, ironically the current fragmentation risks causing audience confusion and consumers to suffer from label blindness. Ultimately, the success of an AI-free label relies on a universal industry definition, transparent regulatory backing, and a rigorous verification system.

    Other authors: Ashley Johnson, Senior Associate

    The information provided is not intended to be a comprehensive review of all developments in the law and practice, or to cover all aspects of those referred to.
    Readers should take legal advice before applying it to specific issues or transactions.