Satyen Dayal, UK Managing Director, Technology, Edelman, explores the growing lack of trust in AI and highlights the barriers to its adoption.

The meteoric ascent of Artificial Intelligence (AI) casts a long, unsettling shadow over a world already riddled with economic unease and societal trepidation.

The 2024 Edelman Trust Barometer Supplemental Report: Insights for the Tech Sector lays bare the raw nerves of a global populace grappling with fears of job loss (88% among the employed), inflation (73%), climate change (76%), hackers (75%), nuclear war (73%), and information warfare (61%) – since last year, fears of hackers and information warfare have surged by five and six percentage points respectively.

In these turbulent times, the trust once placed in AI companies is waning, plummeting from 62% to a mere 54% over five years. Our 2019 optimism for the algorithm now stands on trial in the harsh light of 2024.

How do we restore trust before acceptance of AI becomes futile?

No sheltering from AI

AI companies find themselves on shaky ground with trust in only 12 of 28 surveyed nations, yet the global sentiment is far more nuanced. In regions like Malaysia, the UAE, South Korea, and Kenya, trust has increased by over five percentage points in just a year, with Nigeria seeing an 11 percent leap.

Conversely, in the US, Argentina, Ireland, and Canada, trust has fallen by more than five percent, dropping to a mere 32% in Sweden.

Chart 1: Trust in AI companies

This stands in sharp contrast to the wider tech sector, which despite experiencing fluctuations has seen a cumulative increase in trust across all 28 countries. However, tech is not without its battles, struggling to reclaim its once-dominant number-one position on the global stage over the last three years.

But as the adoption of AI heats up, the disparity between trust in technology companies at large (76% trust) and trust in AI tech (50% trust) is striking – a 26-point chasm underscores a broader disconnect. This divide is also deeply demographic, with the weakest AI tech trust levels among women (47%), those aged 55 and above (39%), and lower income brackets (47%).

Barriers to embracing AI underscore a critical narrative: Job displacement is far from the primary concern. Among those who feel less than enthusiastic about the growing use of AI, their top reason is that AI could compromise their privacy (39%). Second and third are the worries that AI may devalue what it means to be human (36%) or is not adequately tested (35%). Job security is eighth on the list behind AI’s potential harm to society and to people (35% each), that it’s happening too fast (32%), and will worsen inequalities (27%).

This crisis of confidence seeps into deeper anxieties. Notably, only 32% feel in control of AI's impact on their lives and a mere 34% are confident in its regulation.

The dividing lines of innovation

AI stands at the forefront of a global debate on innovation's role in society – one met with trends including vaccine hesitancy, skepticism towards electric vehicles, Hollywood upheavals over AI's encroachment, and the electoral success of pro-fossil fuel agendas.

2024 is a year with elections taking place across more than 64 countries that account for 49% of the global electorate. Innovation emerges as a pivotal political battleground capable of tipping the scales.

Right now, more people think innovation is poorly managed (39%) than well managed (22%). This perceived mismanagement fosters a sense of alienation. Those who feel innovation is mismanaged are 22 percentage points more likely to think that technology is changing too quickly, and 27 percentage points more likely to say society is changing too quickly.

Such disillusionment with the governance of innovation breeds resistance and dampens the embrace of new technologies, placing the acceptance of AI in a precarious balance. Unlike the more warmly received green energy or the vehemently opposed GMOs, AI, akin to gene-based medicine, teeters on the brink, with public sentiment evenly split between acceptance and resistance.

But while in Western democracies resistance to innovation is political (there is far greater resistance to innovation on the right than on the left), AI itself is far less politicized, with little difference in resistance based on political leanings.

Chart 2: Acceptance of Innovation

Transition to the AI economy: Trust in the tech and the human

Across the board, spanning businesses, NGOs, governments, and media, the universal expectation is clear: for AI to be accepted, its creators must listen to and address the public’s concerns, giving them the opportunity to ask questions.

Boosting enthusiasm for AI hinges on a trifecta of support: institutional, expert, and from peers. A significant uptick in AI optimism among those who are less than enthusiastic about it could be achieved if individuals trusted businesses to thoroughly test it (48%); governments to ensure it is implemented safely (46%); media to provide accurate, unbiased analysis (40%); and NGOs to work with local authorities to deal with its impact (38%).

They also would feel more positive if they were to see experts endorse AI (44%) and family and friends embrace it (35%).

Transition to AI is about putting trust in the tech and the humans delivering it. The key to fostering enthusiasm lies in highlighting AI’s societal (53%) and personal benefits (51%), alongside a better understanding of it (51%). Managing the risks means having the confidence that those adversely affected by AI are taken care of (46%), and, slowing down the pace at which it is happening (46%).

Chart 3: Increasing Enthusiasm for AI

For AI to be widely accepted, it must be presented not only as a harbinger of a brighter future, endorsed by the scientific community, but also as a technology that empowers individuals with greater control over its impact.

AI acceptance starts in the workplace

At 79% trust, employers command trust levels among their employees that tower above any of the core institutions.

Among tech employees, trust in AI tech is at 71%. Trust in AI tech is at 73% amongst women in tech, four percent higher than their male counterparts (69%).

Chart 4: Tech Employee Trust in AI

Tech workers have greater optimism for AI than their non-tech counterparts. They are seventeen percentage points more confident in AI regulation than employees in other sectors, and 14 percentage points more convinced that they have a lot of influence over how it impacts their lives.

Tech employees have higher trust in their employers to ensure the introduction of new innovations into society is well-managed (78% compared with 70% for non-tech workers). Executives show the highest trust in institutions at 85%, with heightened expectations for their CEO to engage publicly on ethical technology use (91%), future job skills (89%), and the impact of automation on jobs (88%).

And while 45% of those who work in tech embrace the growing use of AI, 19% reject it. Right now, non-tech workers are as likely to embrace (34%) AI as reject it (32%).

With so much hanging in the balance, the transition to an AI economy must start at work.