Ethics in AI: Building Trust Through Standards and Transparency
Many people feel uncertain or even wary about ethics in AI. Despite this, AI is already behind three times more daily tasks than we think, a new report from the Institution of Engineering & Technology has revealed.
The surge in artificial intelligence (AI) has left a third of us fearing the unknown, yet we have three times as many daily interactions with AI as most people realize, according to the research.
On average, the UK public recognizes AI plays a role in something we do at least once a day – whether in curating a personalized playlist, mapping out the quickest route from A to B, or simply helping to write an email.
However, hidden touch points can be found in search engines (69%), social media (66%), and streaming services (51%), which all discreetly use AI, as well as tools such as Google Translate (31%) and autocorrect and grammar checkers (29%).
Despite its everyday use, more than half of us (51%) admit nervousness about a future with AI – with nearly a third of people feeling anxious about what it could do in the future (31%). Over a third are concerned about privacy (36%) and feeling it will lead to mass unemployment (39%).
Those surveyed who felt nervous do so because of not knowing who controls AI (42%) and the inability to tell what is real or true with AI-generated fakes (40%). They also expressed concerns that AI will become autonomous and out of control (38%). And that it will surpass human intelligence (31%).
But people do recognize and welcome the role it will play in revolutionizing key sectors, such as manufacturing (46%) and healthcare (39%) and specifically medical diagnosis (32%), as well as tackling issues such as climate change (24%).
The UK AI Safety Summit took place at the start of November. It saw global leaders gather to discuss the risks associated with AI and how they can be mitigated through coordinated action.
Research reveals that 36% of Brits think the government needs more to regulate and manage AI development. In comparison, 30% of those nervous about AI feel that government regulations cannot keep pace with AI’s evolution.
Those surveyed also shared their concerns over the lack of information about AI and the lack of skills and confidence to use the technology, with over a quarter of people saying they wished there was more information about how it works and how to use it (29%).
It’s no secret that AI holds the potential to drive innovation and enhance productivity across diverse sectors like construction, energy, healthcare, and manufacturing.
Yet, it’s imperative that we continually evolve ethical frameworks surrounding data and AI applications to ensure their safe and responsible development and utilization.
We need to see greater transparency around the training and operation of AI systems – especially relevant for publicly accessible large language models, like ChatGPT, which trains its models in part on user data, said an AI expert who participated in the summit.
We trust our civil engineers to build bridges that don’t fall down and tunnels that don’t collapse. We can have confidence in them because we know they follow strict internationally agreed standards in their work, added the expert.
We trust our airplanes, our electrical supply systems, and our household electricals for the same reasons. So, we should build software systems, particularly those that make decisions, using a standards-based approach, they concluded.
It’s essential to recognize that AI has a longstanding presence and already forms the foundation of many daily activities, such as facial recognition on social media, navigation on maps, and personalized entertainment recommendations.
We need to see the UK government establishing firm rules on which data can and cannot be used to train AI systems – and ensure this is unbiased. This is necessary to ensure AI is used safely and to help prevent incidents from occurring – and it is fundamental to maintaining public trust, which underpins the economic and social benefits AI can bring.
The IET research on ethics in AI was conducted online by Opinion Matters from 16-18 October 2023 among a panel, resulting in 2,008 nationally representative consumers responding from across the UK.
To learn more about the IET’s work on AI, please visit www.theiet.org/ai-research.
