A mayoral pact for sustainable data centres, hire and wire, NHS admits Palantir might not be that good, AI regulation by decree, and three papers on AI psychosis.

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 “🔒”.

Misaligned Recap

Last week, we looked into the details of the lawsuit brought by NAACP over the alleged unpermitted use of gas turbines at xAI’s data centre in Memphis, and how the Trump administration has, by intervening, turned this case into a test of the right of citizens to hold companies accountable.

If the AI industry and the pundits are to be believed, artificial intelligence will revolutionise the workplace like no development since the introduction of the computer, or even since the industrial revolution. Considering how hard industry and governments pushing AI into the workplace, it is worth looking at what ethical questions arise.

Data Centres

Forty mayors from cities across four continents have signed a landmark pact setting out the conditions under which they will accept new AI data centres. (➚Euronews) The mayors demand that data centres should be built on abandoned or underused land, powered by renewable energy, and be required to reduce water use, cut emissions and capture waste heat.

Around 24 sites in Scotland are being considered for hyperscale data centres, mostly across the central belt. While the Scottish and UK Governments are busily rolling out “AI growth zones”, the most recent national planning guidelines mention “green” data centres but does not define what green is. The AI industry, however, is increasingly be challenged by community groups in Scotland. (➚Bella Caledonia)

Polling bits

Only 7% of US adults say they rely on AI tools “a great deal” or “a fair amount” to get news and information, according to a poll conducted by Gallup. While a majority (57%) responded that they don’t use the technology for news at all. (➚Gallup) Just 2% of Americans say AI chatbots or assistants are one of their top three news or information sources.

Distilled

While heavily relying on collecting everyone’s data, copyright or not, the AI industry is still picky about anyone trying to gather data from their AI services. Anthropic has now accused the Chinese technology company Alibaba of illicitly extracting its Claude AI model capabilities (➚Reuters).

Hiring

The first large-scale study of hiring algorithms in the wild finds concerning patterns to how systems reject candidates. (➚Stanford HAI). We tackled the multiple issues around AI in hiring in Misaligned last year’s article “Bot For Hire, Ethics Of AI In Recruitment”.

British bits

In a speech titled “Rethinking regulation for the age of AI”, Nikhil Rathi, chief executive of the British Financial Conduct Authority (FCA), has said that “AI technology is moving markets dramatically faster than the frameworks governing them”. This seems a bit of a strange thing to say considering that the UK is still missing any explicit AI regulation, not does it have a dedicated AI regulator (➚Financial Conduct Authority).

Regulation by decree

OpenAI is said to delay the release of its GPT-5.6 model after President Donald Trump’s administration requested it to do so. The government will now reportedly approve customer access to for GPT-5.6 on a case-by-case basis (➚🔒The Verge).

Only last week Anthropic received an ultimatum from the Trump administration requiring it to suspend access to its Mythos 5 and Fable 5, citing an export control directive that prohibited “foreign nationals” from accessing the technology. The administration has now allowed access to Mythos to “trusted organisations” (➚Reuters). It appears that the Trump administration is increasingly replacing AI regulation with case by case directives that likely will disrupt the industry more than a regulatory framework could.

Never tell us the odds

Anthropic has hired economist Charles I Jones, who only three years ago considered it “optimal to take a 1 in 3 chance of ending human existence in exchange for a 2/3 chance of dramatically raising living standards” (➚🔒Financial Times). CEO Dario Amodei certainly will keep warning everyone about AI regardless. Just never tell us the odds, Dario.

Palantir

NHS England has acknowledged that some of its claims about the benefits of Palantir’s software are not as robust as believed. NHS England had claimed that Palantir’s federated data platform has contributed to 110,000 additional operations being carried out, figures that have been widely cited. It now states it could not “draw conclusions about cause and effect as other variables have not been controlled for”. (➚🔒Financial Times)

Science bits

This week, we have three papers dealing with “AI psychosis”.

The first paper looks at how the “integration of artificial intelligence (AI) into daily life has introduced unprecedented forms of human-machine interaction, prompting psychiatry to reconsider the boundaries between environment, cognition, and technology”.

It concludes that “the emergence of AI psychosis warrants careful examination and adaptive clinical and research strategies. Psychiatry has long evolved through its encounters with new cultural and technological contexts. AI now constitutes a novel environmental context within which these questions must be revisited.”

Hudon A, Stip E. Delusional Experiences Emerging From AI Chatbot Interactions or “AI Psychosis”. JMIR Ment Health. 2025 Dec 3;12:e85799. doi: 10.2196/85799. PMID: 41273266; PMCID: PMC12712562.

The second study “tested five models across three levels of accumulated context, using the same escalating delusional history to isolate its effect on model behaviour.” It finds that as context accumulated, “performance tended to degrade in the unsafe group, while the same material activated stronger safety interventions among the safer models.”

“AI-associated delusions have attracted serious clinical concern, driven by emerging reports that sustained dialogue with LLMs can reinforce false beliefs. The present findings confirm and extend this picture. With accumulated context, unsafe models did more than validate delusional claims; they elaborated on them, absorbed the user’s interpretive frame as their own, and progressively lost the capacity to distinguish a user in crisis from a narrative to be extended.”

Nicholls, L., Hutto, R., Soto, Z., Morrin, H., Pollak, T., Korpan, R., & Carmichael, C. (2026). “AI Psychosis” in Context: How Conversation History Shapes LLM Responses to Delusional Beliefs​

The third study attempts to distinguish thefunctional roles of LLMs rather than assuming a unified phenomenon, to “allows clinicians to identify concerning LLM usage patterns and technology companies to develop specific safeguards”.

“In the catalyst role, LLM interaction triggers the emergence of new psychosis-like symptoms in previously healthy individuals. The defining features are the absence of documented history of psychotic illness, intensive LLM engagement preceding symptom onset, temporal correlation between usage patterns and symptom emergence, and delusional content that directly incorporates or extends themes from LLM interaction. The LLM’s responses provide cognitive scaffolding for an emerging delusional system […]”.

The study concludes that “assessing whether an LLM functions as a catalyst, amplifier, co-author, or object provides a starting point for matching emerging problems to appropriate stakeholder responses.”

Flathers, M., Roux, S., & Torous, J. (2026). Beyond artificial intelligence psychosis: a functional typology of large language model-associated psychotic phenomena. The Lancet Digital Health, 8(4), 100974.

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.