AI: navigating a moral maze

As AI and large language models continue to evolve, businesses face mounting ethical dilemmas, complicated further by global political shifts. Here, we examine those dilemmas and how businesses can navigate them.

Every day seems to bring a new ethical concern for businesses wishing to navigate their way through the uncharted waters of Artificial Intelligence (AI) and especially the still-new frontier of large language models (LLMs). These are made all the more complicated by political developments across the globe that feed into technological issues. And what are we meant to make of these systems when even the AI companies themselves request you do not use AI when applying for jobs on their careers page?

The latest tremor sent through the world of AI was the launch of another DeepSeek iteration from China – in early tests it seems to perform as well as ChatGPT. Early censorship concerns were quickly raised: it refused to answer questions on Winnie the Pooh and President Xi Jinping, as well as more seriously Tiananmen Square and human rights. Other commentators observed grave privacy concerns compared with most Western tech companies. And yet, the success of China in creating an effective LLM using less energy than existing models (and without reliance on chips previously thought to have been essential) introduced a useful corrective to the Silicon Valley narrative (as well as, of course, to NVIDIA’s stock price).

Any knowledge or creative industry requires protection from theft, which existing copyright laws provide. So the attacks by Sam Altman and others on the apparent copycat nature of DeepSeek certainly sounded a little hollow given the fast and loose approach taken by AI companies in their own approach to copyright. In the United States, The New York Times is leading a group of news organisations in a court case against OpenAI, arguing that the data powering ChatGPT has included millions of copyrighted works from news organisations, amounting to copyright infringement on a huge scale. In the United Kingdom, Sirs Paul McCartney and Elton John among others have voiced opposition to the government’s Data (Use and Access) Bill, which threatens to undermine the copyright of artists and publishers in an attempt to appease the avaricious AI monster’s requirement for scraping data.

Other global plans to try to tether the monster are also looking troubled. The United States and the United Kingdom both refused to sign a communiqué at a French-led summit to ensure that AI development is “open, inclusive, transparent [and] ethical”, making AI “sustainable for people and the planet” and “taking into account international frameworks”. While the UK’s AI Safety Institute is widely recognised as a global leader in attempting to square the circle of best practice AI governance, it sometimes seems that the behemoths of the US West Coast are going to trample on anything that gets in their way. However they will need to respond to business anxieties, in turn reflecting consumer concerns, of trustworthiness – perhaps through a blockchain system for AI model management. This should allow a space for AI innovation while protecting consumer privacy or artist copyright.

What can the individual business do? The first, most obvious action to us is to keep the human in the loop – everything that used to be checked should still be checked and worked on – whether it be figures or editorial content, art or videos. And anybody involved with using these tools will have to satisfy their due process in terms of privacy and ownership of data. On the supply side of the argument, perhaps don’t expect so much from AI – businesses aren’t necessarily going to be able to save vast amounts of money with AI and there are growing concerns around the impact of the use of AI on the climate. It may help with initial ideas and give a direction of travel, but if we still want originality, humour and accuracy in our content, and even more so if we need information to be translated, AI-generated content still requires the intervention of the best professionals.

Discover what Websters can do to support you in perfecting your content for publication – whether human- or AI-generated.