GENAI Content’s Compliance Problem.

You got GPT4 to write copy, MidJourney to create visuals, Eleven Labs to do a voiceover, and Viola! You are all set to roll out a campaign that might win you consumers and industry accolades. But wait, there is one thing – have you checked whether your content ticks the “boxes” on legal, ethical and organisational compliance?

There are known and unknow risks associated with the use of GenAI content in advertising and entertainment. In absence of a framework, many companies are reluctant to adopt Gen AI tools.

Those of you who followed the Writers Guild of America (WGA) Strike resolution, you would have noted that the film companies refused to take responsibility on use of AI. Instead, the buck was passed on to the writers. The resolution concluded that writers can chose any tool they want to work with, as long as they have the consent from the company they have been hired by. Here is hoping that that company will have a policy framework to grant this consent. The companies however WILL NOT demand or direct the use of any AI tools. The resolution literally stated that “a company can’t require the writer to use AI software (e.g., ChatGPT) when performing writing services.” This complex issue has accountability written all over it.

The associated compliance risks are many, and depend on the context as well as the nature of the use. The below list is by no means complete, but gets us on the path to a risk assessment around this issue:

Deceptive Messaging:

Advertising should, by principle, be truthful and not misleading. GenAI content cannot be treated any differently. All content should accurately represent the product or service being advertised. There are regulatory and self-regulatory industry frameworks around this across the world.

Disclosure:

Maintaining transparency is crucial when using GenAI content. Consumers have the right to know if they are interacting with AI-generated material. Advertisers should clearly disclose the use of AI in content creation, ensuring transparency and managing consumer expectations. This is no different from frameworks that now exist around influencer marketing. Whilst a regulation may or may not have mandated disclosure, it only adds to a brand’s credibility if this disclosure is made.

Privacy Compliance:

The use of AI often involves data processing, which may implicate privacy concerns. Did that data have the permission of its owner? The Italian government’s ban on ChatGPT was fairly rooted in this concern. Where it was lacking was in the ability to support the speed of technological development by providing a proactive framework.

Advertisers must comply with relevant data protection laws and ensure that the data used to generate content is obtained and handled in a lawful and ethical manner, with respect to privacy rights. Remember, GDPR compliance is not just about securing your 1P data.

Responsiveness:

All advertising content needs to be up-to-date. There are a lot of AI models that have not been trained on the most up-to-date data. Advertisers’ review of these models prior to use is crucial and can lead to reputation damage if not done properly.

False Endorsements:

Deepfakes are real – pun intended. So is the risk of these deepfakes being used in false advertising. Our world’s capacity to monitor false endorsements isn’t anywhere near the required levels. It is hence important to put in place policies that mitigate legal risks arising out of these. The flip side of the coin might also demand extra diligence to be done on “real celebrity and influencer” endorsement contracts. Did anyone see this Tom Hanks Instagram post a few days ago?

Regulatory Compliance:

Just because a machine is generating it, a piece of advertising does not get any indemnity against local and international laws. In fact, if you read through “terms of service” for many GenAI tool, they pass on the responsibility of regulatory compliance to the user – indemnifying themselves against any regulatory backlash. Are you ready to take on that responsibility?

Industry Self-Regulation:

Like in most cases with technology, it is going to take regulators some time to catch up with the technological development in this area. Whilst that happens, it is important that industry associations and bodies start to rally around self-regulation on this subject.

Intellectual Property / Rights:

Perhaps the most important issue of all. Currently there is NO regulation around the question of “who owns the GenAI generated content”. There is a na?ve assumption of “whoever prompts, owns” around this issue among many users. Any content that is generated on the back of data sets that are owned by a wide number of sources is a regulatory nightmare. It is for this reason that a recent GenAI tool by “Getty Images” has been developed using its own set of licensed images.

Whilst your organisation may or may not have taken a step towards risk assessment and mitigation on GenAI content, we at So&So have dug deep into this issue and have partnered with legal and compliance experts to develop an approach that can get you started. Reach out to us.