Dominic Rowles, lead ESG analyst, Hargreaves Lansdown: “AI has the potential to be a powerful tool that enhances our lives by automating tasks, personalising things like financial services, education and healthcare, and increasing our understanding of the world around us, and create exciting investment opportunities as it grows. But it could also make privacy a thing of the past, threaten international relations and reinforce discrimination. Shareholders in AI companies should be aware of the five key ethical risks associated with the technology before making an informed investment decision.
1. Deepfakes and fake news
Deepfakes use AI to create realistic but fake audio or video recordings, aften altering existing footage to misrepresent someone’s words or actions. These clips can be used to spread misinformation, create fake news, or damage reputations.
As deepfakes become easier to create and more convincing, they could undermine public trust in genuine media, causing people to doubt real news and information. This growing scepticism could help misinformation flourish, making it harder for society to have productive discussions. Worse still, if a deepfake misrepresents political leaders or events, it could incite tensions and undermine trust between nations, potentially even leading to conflict.
2. Mass unemployment
AI systems and robotics can carry out increasingly complex tasks more quickly and cost effectively than human workers. AI-powered robots are already transforming factories worldwide. Tools that can monitor and analyse production line processes in real-time, predict equipment failures and use historical data to suggest solutions, are becoming common on factory floors.
Many other jobs could face being displaced by AI in the not-too-distant future as well, including customer service representatives, computer programmers, graphic designers and journalists. As AI develops, it’s likely to claim more and more jobs over time. A report released earlier this year suggested that in a worst-case scenario, as many as 8 million jobs could be lost to AI in the UK alone, with roles traditionally carried out by younger workers and those on lower wages most at risk.
3. Privacy and Surveillance
AI can process data at extraordinary speeds, accomplishing in seconds what would take human analysts years. In the wrong hands, AI could be used to collect personal data without consent, leading to serious privacy violations and misuse of information. Governments and companies could also use AI-powered tools to monitor people, suppress personal expression and influence the way they make purchasing decisions, or even the way they vote.
4. Discrimination
AI systems learn by being fed vast amounts of data, and if that data reflects existing society-wide race, gender or socioeconomic biases, the algorithm will carry those biases too. If an AI system is used in hiring, for instance, it could discriminate against particular groups by favouring CV’s that resemble those of historically successful candidates.
5. Accountability
When AI systems make decisions independently, it can be difficult to work out who is responsible when they malfunction, or produce unexpected results – whether it’s the developers, the users, or the company that deployed the system. At the same time, AI is advancing at a pace that outstrips most people's understanding, and global regulators are scrambling to keep up. That means that issues of AI liability are complex, and often not covered by an existing legal framework.
Balancing innovation and responsibility with opportunities
In the absence of overarching AI regulation, ensuing companies uphold ethical practices is crucial. Investors have an important part to play, both by pushing AI companies to adopt rigorous ethical standards and advocating for policymakers to establish a regulatory framework that promotes responsible AI development. As with any investment, shareholders will have to balance the exciting growth opportunities in this fast-emerging sector, with the risks associated with the new technologies.
Our equity analysts think that the best opportunities currently sit outside just the small cohort that grab headlines. For opportunities further afield that could add some much-needed diversification, see our link to three shares that could benefit from the AI transition.
|
