As artificial intelligence becomes more deeply embedded in business, infrastructure and everyday life, public resistance is growing alongside adoption. Anthropic CEO Dario Amodei believes the reason goes deeper than concerns about job losses, data centers or increasingly capable AI systems.

In a recent exchange with investor Gavin Baker, Amodei argued that the backlash surrounding AI is fundamentally a crisis of trust between the public and the institutions developing and deploying the technology.

The debate raises an important question for the entire technology industry: can AI earn widespread acceptance through better communication, or will companies ultimately have to prove its value through measurable outcomes?

The debate over AI’s increasingly negative image

Baker argued that warnings from prominent AI executives about the potential dangers of advanced systems have contributed to growing skepticism toward the industry in the United States.

He suggested that Amodei, who has repeatedly discussed risks associated with advanced AI, should become a more positive advocate for the technology and its potential benefits.

Amodei rejected the idea that his public position has been disproportionately negative.

He pointed to his previous writing about AI’s potential to accelerate scientific discovery, improve healthcare and transform economic productivity. His essay Machines of Loving Grace, for example, was intended to describe a more ambitious and optimistic vision of what powerful AI systems could accomplish.

But Amodei agreed on one important point: the public perception problem is real.

Where he disagrees is on its cause.

AI has a trust problem

According to Amodei, public skepticism toward AI cannot simply be blamed on executives discussing its risks.

Instead, it reflects a much broader erosion of confidence in corporations, governments and the technology industry that has developed over decades.

AI is simply the latest technology to inherit that distrust.

This distinction matters.

If the industry’s problem were primarily one of communication, companies could potentially address it through better messaging, education and marketing.

If the underlying problem is trust, however, the solution becomes considerably more difficult.

AI companies need to demonstrate that the technology produces meaningful benefits, that risks are being managed responsibly and that the enormous power associated with advanced AI systems will not remain concentrated among a small number of organizations.

Promises are no longer enough

Amodei also offered a significant criticism of his own industry.

AI companies, including Anthropic, have made ambitious promises about how artificial intelligence could improve society. But many of those promises have yet to translate into outcomes that ordinary people can experience directly.

Talking about AI potentially curing diseases, accelerating scientific research or dramatically increasing productivity can only go so far.

The industry eventually has to deliver.

That represents a shift in the AI conversation. Over the past several years, much of the industry’s momentum has been driven by capability demonstrations: increasingly sophisticated models, larger context windows, multimodal systems and autonomous agents.

The next stage may be judged differently.

Organizations will increasingly need to demonstrate real-world value rather than technological potential.

Regulation and the concentration of AI power

The discussion also touched on one of the industry’s most contested questions: whether AI regulation protects society or strengthens the position of the largest technology companies.

One common Silicon Valley argument is that regulation creates compliance costs that large AI companies can absorb more easily than startups, ultimately consolidating the market around a small number of powerful players.

Amodei argues that this presents a false choice.

In his view, carefully designed regulation could simultaneously address serious AI risks and constrain the power of frontier AI companies while leaving space for smaller competitors and open-weight models.

This is particularly important because advanced AI has structural characteristics that can encourage concentration.

Training and operating frontier models requires enormous computing resources, specialized chips, infrastructure, capital and technical expertise. Open-weight models can distribute access more widely, but they do not completely eliminate the underlying dependence on compute infrastructure.

The challenge for policymakers is therefore not simply deciding whether AI should be regulated.

It is designing rules of the road that improve safety without unintentionally locking the market around today’s largest players.

What this means for technology leaders

For CTOs, CIOs and companies implementing AI, the trust debate is no longer something that concerns only frontier model developers.

Every organization deploying AI inherits part of the responsibility.

Businesses need to be able to explain where AI is being used, what data it can access, how decisions are validated and where human oversight remains necessary.

They also need to demonstrate why the technology is being introduced in the first place.

An AI implementation that reduces repetitive work, improves customer support, helps employees find information faster or makes an operational process more reliable provides a much stronger argument for adoption than another abstract promise about transformation.

Trust will increasingly become part of AI architecture itself.

Governance, observability, security, transparency and human oversight will need to sit alongside model performance when organizations evaluate AI systems.

From AI hype to AI proof

The technology industry has spent years explaining what artificial intelligence could do.

The next phase will be about proving what it does.

For companies building AI products or integrating AI into existing systems, that creates a useful strategic principle: start with measurable problems, not AI capabilities.

Identify where the technology can produce tangible improvements. Build appropriate safeguards around it. Measure the outcome. Make the system understandable to the people who use it.

The organizations that succeed with AI may not be those making the biggest promises.

They may be the ones that give customers, employees and partners the strongest reasons to trust what they build.

Control F5 Software perspective

At Control F5 Software, we believe successful AI adoption starts with the business problem, not the model.

Whether AI is introduced into customer support, internal knowledge systems, operational software or complex enterprise workflows, the objective should be the same: build technology that creates measurable value while remaining secure, understandable and aligned with the way the organization actually works.

Because as AI moves from experimentation into critical business processes, technical capability alone will not determine adoption.

Trust will.

Source