Every chart this week pointed up. SHRM told a room of 20,000 HR pros in Orlando that 92% of CHROs want deeper AI. Investors handed $60 billion to a single coding startup. Chinese labs shipped open models and closed mega-rounds. But the people actually typing into these tools just told Pew something different. Nearly half of US adults now use AI chatbots, and only 16% think AI will make society better. That split is the AI adoption trust gap. It is the number HR leaders should track this week, not the adoption rate everyone keeps quoting.

The AI Adoption Trust Gap Just Got a Number

Pew Research Center released a national survey on June 17 that lands awkwardly against the week’s optimism. About 49% of US adults now use AI chatbots like ChatGPT, Gemini, or Copilot. That is up from roughly a third in 2024, and ChatGPT leads at 44%. (Source: Pew Research Center)

Then comes the uncomfortable part. Only 16% expect AI to benefit society over the next 20 years, while 40% expect harm. In addition, 63% say AI is moving too fast, and 71% worry about their personal data. Worse, 67% have little or no confidence that government can regulate it. (Source: TechCrunch)

Why the trust gap matters for HR leaders

If you run people ops, your rollout plan is probably built backwards. The usual assumption treats adoption as the hard part and trust as something that follows. However, Pew flips that order. People will use the tool every day and still distrust it. So your AI screening, benefits chatbot, and auto-drafted performance notes all land on employees who already expect AI to slip.

For a 200-person company, the resistance you hit is not about whether people can use AI-driven hiring tools. It is about whether they believe those tools are fair. A resume screen nobody trusts creates more grievance work, not less.

What to do: Before you expand any AI tool that touches employees, publish how it works. Spell out where a human stays in the loop. Trust is earned through disclosure, not through a slicker demo.

92% of CHROs Want More AI, but Adoption Is Stuck at 39%

At SHRM26 in Orlando this week, June 16 to 19 with roughly 20,000 attendees, “AI, Data and Tech” was the standout track. SHRM’s own figures tell the same split story. 92% of CHROs anticipate deeper AI integration, yet only 39% of HR organizations have adopted it. (Source: SHRM)

That 53-point gap between intent and reality is the AI adoption trust gap inside your own building. Leaders want it; teams have not shipped it. For most HR leads, the bottleneck is rarely budget. Instead, it is proof that the tool is accurate and safe enough to put in front of staff. The same caution shows up in the wider AI skills gap in HR. So start with one low-stakes workflow, measure it, then earn the next one.

A $60 Billion Acquisition While the Public Pumps the Brakes

SpaceX disclosed an all-stock deal to buy Anysphere, the maker of AI coding tool Cursor, for $60 billion on June 16. (Source: TechCrunch) It is the largest acquisition of a venture-backed startup on record, and it landed four days after SpaceX’s own $75 billion IPO. Cursor’s revenue run-rate sits near $4 billion.

So capital is sprinting in the exact week the public asked AI to slow down. For founders, that contrast is the real signal. Money is not waiting for trust to catch up. As a result, if you buy or build AI tools, you are now ahead of your own employees’ comfort level. That makes change management the part you cannot skip.

When a Bank Cuts Off an AI Model, Trust Becomes Policy

JPMorgan removed Anthropic’s Claude from the approved-model list for its Hong Kong staff, following a similar Goldman Sachs move in April. (Source: American Banker) The trigger was not performance. It was contract wording, because Anthropic restricts use of its models across Greater China.

This is the trust gap turning into procurement rules. Big employers are now blocking specific models by geography and terms of service, not by capability. Meanwhile, if your company operates across borders, your AI vendor’s fine print is becoming an HR and compliance question, not just an IT one.

Quick Hits

• DeepSeek closed its first outside round, about $7.4 billion at a valuation above $50 billion. But it routed the money into a vehicle controlled by founder Liang Wenfeng, with almost no investor voting rights. (Source: CNBC)
• OpenAI’s new LifeSciBench graded AI on 750 real research tasks, built with 173 PhD scientists. The best model cleared just 36%. (Source: OpenAI) Another reminder that “AI does research” is still mostly aspiration.
• Zhipu released GLM-5.2 under an MIT license, an open-weight model that rivals top coders at a fraction of the cost. (Source: The Decoder) Open Chinese models keep undercutting closed ones.

If this week has you rethinking where AI touches your people, the fix starts with systems your team can see into. Asanify’s payroll and HR platform keeps a human approval step on AI agents for HR workflows. So adoption never outruns trust. Worth a look if you are scaling AI in HR this year.

AI Adoption Trust Gap: Common Questions

What is the AI adoption trust gap?

It is the split between how many people use AI and how few trust it. Pew’s June 2026 survey found 49% of US adults use AI chatbots, yet only 16% believe AI will benefit society. High usage sits next to low confidence.

Why should HR leaders care about AI trust, not just adoption?

Employees often use AI daily and still doubt it is fair or accurate. So if your AI tools for hiring or payroll lack transparency, low trust creates disputes. It slows the rollout, even when the technology works.

How can companies close the AI trust gap with employees?

Disclose how each AI tool works, keep a human in the loop on decisions that affect people, and start with low-stakes workflows. Trust grows from transparency and proven results, not from a polished demo.

Not to be considered as tax, legal, financial or HR advice. Regulations change over time so please consult a lawyer, accountant  or Labour Law  expert for specific guidance.