Why AI Trust — Not AI Capability — Is the Real Bottleneck in 2026

The Quiet Disappearance of Half a Million Coding Jobs

Federal Reserve researchers, using an expanded definition of software workers that includes contractors and coders outside the tech sector, found roughly half a million fewer coding jobs exist today than pre-AI trends would have predicted.

The job market isn't in freefall. But the floor is quietly shifting underneath it — and it's the people just starting out who are feeling it first. Entry-level positions, junior developer roles, and contractor gigs are thinning out in ways that don't make headlines but reshape career paths.

"AI Can Only Diffuse at the Speed of Trust"

Anthropic CEO Dario Amodei keeps returning to a single phrase: AI can only "diffuse at the speed of trust." And right now, trust is running short.

Part of the problem, he argues, is that AI companies — his own included — haven't fully delivered the promised benefits yet. The productivity gains are real in pockets, particularly in software development and certain knowledge work, but they haven't translated into broadly visible improvements in living standards.

"Is that just propaganda? Is that just vapourware that's not going to happen? We actually have to make it happen." — Dario Amodei, CEO of Anthropic

That's a more humble posture than most AI CEOs tend to adopt publicly. It also sets up an uncomfortable tension.

The Numbers That Define the Trust Gap

Claude Mythos: Capability Without Consensus

Anthropic this year unveiled Claude Mythos — a model that cybersecurity researchers say represents a genuine step change in automated vulnerability discovery. It is capable of finding and exploiting zero-day flaws in major operating systems without human guidance.

The company's AI coding tools have helped trigger a wave of stock selloffs in the software sector. Nearly $3 trillion in market cap has evaporated since October, as investors worry AI agents are quietly eating the application layer.

Whether that counts as "delivering the benefits" depends a lot on who you ask — and whose job is next.

The Policy Gap: Same Words, No Mechanisms

Not everyone is convinced the industry's newfound policy enthusiasm is the real thing. Soribel Feliz, a former U.S. Senate AI policy advisor, noted that the ideas circulating — shared prosperity, democratized access, worker protections — are essentially the same framework discussed since ChatGPT launched in late 2022.

"I have it in my handwritten notes. All of this was already said, all of it." — Soribel Feliz, former U.S. Senate AI policy advisor

The problem, she said, isn't the vocabulary. It's that nobody has built the actual mechanisms to back it up.

Economists Push Back on Industry Timelines

Economists who have spent careers studying technological transitions are also skeptical of the timeline predictions coming from AI executives. Daron Acemoglu, Erik Brynjolfsson, and David Autor — some of the most cited labor economists working on this question — have pushed back on sweeping job-loss projections.

Their argument: the evidence so far points to slower, more uneven disruption than the industry tends to advertise. Technology adoption rarely follows the exponential curves that Silicon Valley pitches to investors.

Amodei's Regulatory Vision: Cars and Planes, Not Bans

Amodei wants AI regulation modeled on how society handles cars and planes — acknowledging economic value while enforcing real safety standards. His actions so far:

• $20 million behind a PAC pushing for stricter AI safety laws
• Clashed with his own government over attempts to restrict Anthropic's Pentagon contracts
• Publicly acknowledged that AI companies haven't yet delivered on their promises

Whether any of that adds up to meaningful protection for the entry-level analyst staring down an uncertain career path is still an open question.

What This Means for Workers and Businesses

The takeaway isn't that AI is going to replace everyone overnight. It's that the disruption is already happening unevenly, and the people with the least leverage are absorbing the most impact:

1. Entry-level roles are shrinking first. Junior developers, analysts, and contractors face the steepest decline in opportunities.
2. Productivity gains are real but narrow. Software development and knowledge work see measurable improvements; most industries don't — yet.
3. Trust is the adoption bottleneck. Until AI delivers broadly visible benefits, enterprise adoption will remain cautious.
4. Policy is all talk, no infrastructure. The frameworks exist on paper. The enforcement mechanisms don't.
5. The market is repricing risk. $3 trillion in software market cap losses reflect investor uncertainty about which companies survive the AI transition.

Frequently Asked Questions

How many coding jobs has AI eliminated?

Federal Reserve researchers found roughly half a million fewer coding jobs exist today than pre-AI trends would have predicted, using an expanded definition that includes contractors and coders outside the tech sector.

What does "AI diffuses at the speed of trust" mean?

Anthropic CEO Dario Amodei argues that AI adoption is limited not by technical capability but by public and institutional trust. Until AI delivers broadly visible improvements in living standards, skepticism remains rational.

Are executives seeing AI productivity gains?

A survey of over 1,000 executives found that 80% said AI wasn't having any measurable impact on their productivity or headcount, despite pockets of real gains in software development and knowledge work.

What is Claude Mythos?

Claude Mythos is an Anthropic AI model that cybersecurity researchers say represents a step change in automated vulnerability discovery, capable of finding and exploiting zero-day flaws in major operating systems without human guidance.

What do economists say about AI job-loss predictions?

Leading labor economists Daron Acemoglu, Erik Brynjolfsson, and David Autor have pushed back on sweeping job-loss projections, arguing that evidence points to slower, more uneven disruption than the AI industry tends to advertise.