spot_img

Date:

Share:

Who owns an AI’s “masterpiece”? – The South African reality

Artificial Intelligence (AI) and the way that it can be used is evolving at a rapid rate. Many South African businesses, including startups and SMME’s, already use AI tools, whether for process automation, increasing efficiency, or obtaining data-driven insights.

Most of us are familiar with generative AI, which creates outputs (like text, images and code) in response to prompts typically made by humans – think ChatGPT and DALL-E.

The step up from this is Agentic AI. Agentic AI builds on generative AI and is more proactive and autonomous, using human-esque reasoning to achieve specific goals and create outputs with minimal human input – think AutoGPT or Copilot agents.

So, who owns these outputs? 

South African copyright law is primarily governed by the Copyright Act 98 of 1978 (the Act).  Despite the age of the Act and the rapid advancements of technology over time, the Act gives us a degree of legal certainty regarding the ownership of AI outputs.

This is because the Act makes provision for computer generated works. Computer generated literary, artistic, musical and dramatic works are owned by the person who undertakes the arrangements necessary for the creation of the work.

Is the person who makes the necessary arrangements the AI system programmer? Or is it the person who created the prompt to generate an output from an AI system? Based on the Act, the AI system programmer is likely to be the author and the owner of any AI generated outputs, provided that the AI output meets the other requirements of the Act.  This may, however, not always be the case.

Even though the terms of use of most AI platforms state that users own the copyright in AI generated outputs, these terms of use cannot override what is provided for under South African law. This discrepancy creates uncertainty when using AI generate outputs in reality.

Using AI generated content in your business, without first confirming factual ownership of the copyright poses several risks, including reputational, infringement, indemnity and liability risk. To mitigate these risks, businesses should implement robust AI governance and IP policies, tailor contractual provisions relating to indemnity and liability capsl and manage third party infringement claims.

Carla is a partner in the firm’s TMT and IP practice group, and a member of the firm’s AI Legal Advisory Working Group

spot_img
spot_img

━ More like this

AI in stockbroking: why co-pilot must come before autopilot

As financial institutions race to deploy AI agents, the debate is quickly shifting from what artificial intelligence (AI) can automate to what it should...

Google Cloud Summit in Africa highlights the Continent’s digital transformation and unveils new Agentic AI and Infrastructure Investments

Google announced five major AI initiatives, spanning research, digital skilling, startup funding, and infrastructure Google Cloud today hosted its inaugural Cloud Summit in Africa at...

Insurers grapple with new fraud threat: AI-generated images

Fake or altered photos depict evidence, from crash scenes to receipts, to support fraudulent claims. Insurance fraud, unfortunately, is inevitable. A certain amount of fraud is...

The important role of assurance in unlocking the full business potential of AI

Artificial intelligence is rapidly moving from experimentation to everyday business practice. Across South Africa, organisations are using AI to improve productivity, automate routine tasks,...

Engineering for density – why liquid cooling is no longer optional for data centres in the age of AI

As enterprises accelerate their adoption of Artificial Intelligence (AI) and High‑Performance Computing (HPC), many are discovering that the physics of the data centre are...
spot_img