Time Equals Trust

Falling Trust, Rising Dependence

One of the lessons from the Professional Speakers Academy is that the more you charge, the more time you must spend with people. Time equates to trust. The higher your price, the longer it takes to persuade people to work with you.

The same holds true in enterprise sales. The most successful sales people cultivate multiple relationships and spend considerable time with decision makers. The secret to higher conversions is time spent after the presentation of a product. This is when you provide reassurance and address any lingering doubts.

Trust in technology increases with familiarity. The network effect of email or social media means the technology gets more useful with time. Confidence increases along with certainty about effectiveness.

All of which makes AI stand out. The KPMG Trust in AI Report 2025 reveals that AI is considered less reliable, even as usage continues to grow. The more people use AI, the more they experience its flaws and yet they become dependent on it.

AI as Normal Technology

Narayanan and Kapoor argue that AI is a normal technology, as opposed to being superintelligence. As such, it diffuses through society in a standard way.

Diffusion is different from both capability and adoption. Models may continue to improve at pace. They may be rolled out to businesses and consumers. But diffusion means that they are used for practical and productive ends. The barriers to that are human.

We see this in slow automation of call centres despite AI being one-tenth the cost of humans. We see it in gradual adoption of driverless cars, even though they are already five times safer than human drivers. OpenAI realised this when usage of reasoning models was low single digits, months after being rolled out to customers.

One of the consequences of launching GPT-5 is increased use of reasoning, as AI chooses how to answer queries. This increases OpenAI’s costs, with the expectation that greater time spent with ChatGPT will increase trust in its output.

More Usage, Less Trust

This move may be in the nick of time. The KPMG survey revealed that while over 70% of people accept AI, only 46% say they fully trust it. Both trust and willingness to rely on AI have fallen since 2022, while worrying about it rose from 49% to 62% of users.

There has been an explosion in use of AI. Later adopters are less tech-savvy and less inclined to trust innovation. They are exposed to constant scepticism and fearmongering in the media. It is little surprise that trust is falling.

KPMG offers other explanations. Increased use and awareness expose people to hallucinations, examples of bias and ethical concerns. The firm calls this a healthy recalibration, reflecting more realistic expectations and reduced hype.

Yet usage increases despite this. Half of employees say they are sometimes or often unable to complete work without AI. They recognise the benefits of speed, personalisation and access to information. Half of employees also fear being left behind if they do not use AI.

The report concludes that AI both enables growth, fairness and innovation, and increases the risks of bias, misinformation and deskilling. People find it useful and unavoidable, while at the same time being aware of its flaws and risks.

Time Still Equals Trust

Narayanan and Kapoor address the rapid adoption of AI in terms of network effects. The internet and social media were slow to get going and then accelerated as more people became users. AI is a tool for individuals. Its pace of adoption may slow, especially if trust in its outputs declines.

The authors are not AI sceptics. They are sceptical of the claims humans make for AI. The reason is bitter experience of the rollout of previous innovations. Most individuals resist change because it takes them outside of their comfort zone. Changing an enterprise or institution takes considerable time, given the layers of human sceptics increasing inertia.

It may be that AI was greeted with enthusiasm by the technically-minded. Now surveys reflect a more balanced view given the larger number of users. The fact that people continue to use AI even when sceptical, bodes well for long-term diffusion.

The teaching of the Professional Speakers Academy still holds. The speed with which AI came onto the scene was deceptive. Only now are we settling into the more time, more trust relationship cycle. As we figure out what AI can and cannot do for us, our reliance on it will build in the areas where we have greatest confidence.

Questions to Ask and Answer

1. How many of my team are using AI at work?

2. How many of them are using personal AI subscriptions?

3. How many of them have had training in secure AI adoption?

The KPMG survey revealed:

• 44% of businesses lack clear governance for AI systems

• Just 37% audit AI for unintended outcomes

• Yet 61% of leaders say AI is critical to their future strategy.

 If you are concerned about AI adoption, find out more about Workforce AI Training.