Most people do not think AI will make things better, too many ask AI about health, and AI can’t cope with users distrusting it.

Welcome to a new edition of misaligned bits, the (roughly weekly) newsletter from Misaligned where we sum up recent news and research, sometimes with a lighter touch.

As usual, we will mark all non-medium links with “➚” (external link) and all possibly paywalled links with “🔒”.

Polling bits

A survey of 350 global business executives with a revenue of at least $1 billion by Gartner has found that many have reduced their workforce irrespective of AI adoption (➚Fortune).

Meanwhile, a survey conducted by the University of Pennsylvania finds that only 17% of Americans believe AI will have a positive impact on the United States over the next decade, while 42% expect its effects to be negative. (➚Eurekalert)

In the UK, one in seven people are using AI chatbots for health advice instead of seeing their GP, a study has found (➚The Guardian). A fifth of respondents said AI chatbots did not encourage them to seek a professional opinion. A similar proportion said they decided against seeking a consultation because of something an AI chatbot had told them. In January OpenAI reported that ChatGPT answers about 40 million health care related prompts every day (➚OpenAI).

Regulatory Bits

The European Commission has published its draft guidelines on the classification of high-risk AI systems. The guidelines aim to support providers and deployers in assessing whether an AI system should be classified as high-risk (➚European Commission). Stakeholders, including providers and developers of AI systems, businesses and public authorities as well as academia, research institutions and citizens, are invited to share their views by 23 June 2026 (➚European Commission).

Data centre bits

Nearly 50,000 Lake Tahoe residents have to find a new power source after their energy source looks to redirect lines to data centres: NV Energy, the Nevada utility that has supplied the bulk of Lake Tahoe’s electricity for decades, told the company that services the region that it will stop providing power after May 2027 because NV Energy needs the capacity for data centres (➚Fortune).

Copyright bits

A federal judge has asked lawyers for more information on Anthropic’s proposed $1.5 billion settlement with authors ​who accused it of misusing their books to train its AI ‌chatbot Claude. The district judge did not grant final approval at the hearing in San Francisco (➚Reuters).

Legal Bits

Yet another lawsuits has been filed against OpenAI for the death of a ChatGPT user, this time related to drug advice. The parents of a 19-year-old who died of a drug overdoes last year allege that the man was coached ‌to take a dangerous combination of substances by ChatGPT. (➚Reuters)

Rather embarrassing

EY (Ernst & Young) has withdrawn a study on loyalty rewards programmes that included apparent AI hallucinations and fake footnotes. The study, which was used by EY consultants in Canada to market their cybersecurity business, used made-up data, mis-attributed citations and referenced a McKinsey report that does not exist. (➚🔒Financial Times)

Science Bits

Researchers found that distrust in AI causes people to give less detailed symptom reports, potentially reducing the accuracy of digital healthcare assessments. (➚Nature)

“Our findings indicate a bias in how users communicate symptoms in digital settings. This outcome could compromise the performance of consumer-facing AI tools in real-world applications, regardless of the underlying model’s actual capacity.”

Reis, M., Reis, F., Kim, Y.J. et al. Reduced symptom reporting quality during human–chatbot versus human–physician interactions. Nat. Health (2026). https://doi.org/10.1038/s44360-026-00116-y​

Follow and subscribe

If you stumbled across this article on the web, subscribe to the “misaligned bits” newsletter and follow Misaligned on Medium, LinkedIn, Threads or Mastodon.

You can now also subscribe to this newsletter directly with your email.