GPT-4o: In summary, AI makers need to act swiftly by engaging with rightsholders, securing licenses, and complying with copyright laws.

Question 1 of 2 for ChatGPT-4o: Please read the post by law firm Wiggin called “Generative AI: Society of Authors warns of unauthorised use of copyright-protected works” as well as the full letter of the Society and tell me what the key messages are.

ChatGPT-4o’s answer:
The key messages from both the Wiggin LLP article and the Society of Authors (SoA) letter focus on concerns regarding the unauthorized use of copyright-protected works by AI developers.

1. Copyright Infringement: The SoA emphasizes that using its members' works—writers, illustrators, scriptwriters, and translators—to train or develop AI models without permission constitutes copyright infringement under the UK Copyright, Designs and Patents Act 1988 (CDPA). The relevant provisions include Section 16, which grants authors exclusive rights to their works, and Section 17, which prohibits unauthorized copying.

2. Licensing and Consent: The SoA strongly urges AI developers to agree to commercial licensing terms before using copyrighted works. It stresses that licensing opportunities are available and should be pursued by AI developers to ensure legal compliance.

3. Unauthorized Use Not Permitted: The letter clearly states that SoA members do not authorize the use of their works for any aspect of AI development (including training and operation) without specific agreements in place.

4. Demands to AI Developers: The letter outlines specific steps that AI developers should take to remedy unauthorized use:

   • Identify the works used to develop AI models.

   • Obtain permission for future use of SoA members' works.

   • Ensure proper remuneration for past and future use.

   • Provide attribution to the original authors where applicable.

   • Remove any works used without permission from their systems and provide evidence of compliance.

5. Opt-out and Further Action: The Wiggin LLP article notes that the letter could be seen as an opt-out under the EU's DSM Copyright Directive, which allows rightsholders to exclude their works from data mining without consent. It also reinforces that legal remedies for SoA members are reserved if developers fail to comply.

6. Balancing AI and Creativity: The SoA’s position is that collaboration between creators and AI systems is essential for the success of AI technologies, which rely on human-created content. Licensing agreements would ensure that both the AI industry and creators can thrive together.

These points reflect the growing concern within creative industries over the unregulated use of copyrighted material in AI development. The SoA is taking action to protect its members' rights and calling on AI developers to respect intellectual property laws.

Question 2 of 2 for ChatGPT-4o: What should AI makers be doing in response to this and what can happen if they don't do anything?

ChatGPT-4o’s answer:
AI developers should take several key actions in response to the Society of Authors' (SoA) letter and similar warnings from creative industries. If they fail to do so, they could face significant legal and reputational consequences. Here's what they should do, and the potential risks if they ignore these concerns:

What AI Makers Should Be Doing:

1. Identify and Disclose Usage of Copyrighted Works:

   • AI makers need to identify which copyrighted works they have used to train or develop their AI models. This transparency is crucial to understanding the scope of any potential infringement.

2. Obtain Proper Licenses:

   • They should engage in licensing discussions with authors, illustrators, and other rightsholders before using their works for training models or generating content. This involves negotiating fair compensation for the use of protected material, as the SoA suggests licensing opportunities exist.

3. Seek Permission for Future Uses:

   • Before using any copyrighted material in the future, AI developers must obtain explicit consent from rightsholders. This ensures that future developments comply with copyright laws and ethical standards.

4. Provide Fair Remuneration and Attribution:

   • AI developers must commit to paying appropriate remuneration to rightsholders for past and future use of their works. Additionally, they should ensure that creators receive proper attribution where their work influences AI outputs.

5. Remove Unauthorized Content:

   • If any copyrighted works have been used without permission, AI developers need to remove them from their systems and provide evidence of compliance to avoid legal repercussions.

6. Maintain Open Communication:

   • AI companies should respond to SoA's letter within the requested timeline (e.g., 21 days for substantive responses) to avoid escalation. Failing to engage with rightsholders could lead to further disputes and legal actions.

What Can Happen If AI Makers Don't Comply:

1. Legal Action and Copyright Infringement Claims:

   • If AI developers continue to use copyright-protected works without authorization, they could face lawsuits for copyright infringement. Under the UK Copyright, Designs and Patents Act (CDPA), penalties can include significant financial damages and injunctions that stop the use of infringing AI models.

   • In the EU, ignoring the opt-out provisions of the DSM Copyright Directive could similarly result in legal challenges.

2. Fines and Financial Penalties:

   • Courts can order damages for both past and ongoing use of copyrighted materials. The SoA and its members could demand compensation for any unlawful exploitation of their work.

3. Reputation Damage:

   • Failing to address these concerns could seriously harm an AI company’s reputation. Creators and other industries might see the company as exploitative or unethical, potentially leading to boycotts, bad press, and loss of trust with consumers or partners.

4. Operational Disruption:

   • If a court rules against AI developers, they may be forced to cease certain operations, retrain their models without copyrighted material, or even face injunctions that limit their ability to develop new AI systems. This could halt progress and cause delays in product development.

5. Missed Opportunities for Collaboration:

   • By not engaging in licensing agreements, AI developers could miss out on beneficial partnerships with creators. Collaborations could lead to a more sustainable, ethical use of AI, but failure to collaborate might isolate developers from key industries.

In summary, AI makers need to act swiftly by engaging with rightsholders, securing licenses, and complying with copyright laws. Failure to address these issues could result in costly legal battles, reputational damage, and significant operational challenges.