A simple AI model has been shown to perform on a par with experienced dermatologists when assessing the aggressiveness of a common form of skin cancer, squamous cell carcinoma. The research was headed by the University of Gothenburg.

Each year, more than 10,000 Swedes develop squamous cell carcinoma. This is the second most common form of skin cancer in Sweden, after basal cell carcinoma, and its prevalence is increasing rapidly. Squamous cell carcinoma often develops in the head and neck region and other areas exposed to the sun over many years.

“This type of cancer, which is a result of mutations of the most common cell type in the top layer of the skin, is strongly linked to accumulated UV radiation over time. It develops in sun-exposed areas, often on skin already showing signs of sun damage, with rough scaly patches, uneven pigmentation, and decreased elasticity,” says associate professor and dermatologist Sam Polesie, who led the study.

Squamous cell carcinoma diagnosis is often easy – the challenge lies in the preoperative assessment – determining how aggressively the tumor is growing to plan and prioritize surgery appropriately. If the tumor is more aggressive, the surgery needs to be scheduled promptly, with more adjacent tissue removed. For less aggressive tumors, narrower margins can be used, with simpler procedures sufficient in some cases.

Almost identical performance

In many countries, Sweden included, preoperative punch biopsies are not routinely performed for suspected squamous cell carcinoma. Surgery is instead carried out based solely on the clinical suspicion of a tumor, with the entire excised specimen sent for histopathological analysis. The fact that surgery is performed without a preoperative biopsy underscores the need for assessment alternatives that do not require tissue samples, such as image analysis using artificial intelligence (AI).

For the study, the researchers trained an AI system in image analysis using 1,829 clinical close-up images of confirmed squamous cell carcinoma. The AI model’s ability to distinguish three levels of tumor aggressiveness was then tested on 300 images and compared with the assessments of seven independent experienced dermatologists.

The results, published in the Journal of the American Academy of Dermatology International, show that the AI model performed almost identically to the team of medical experts. At the same time, agreement between individual dermatologist assessments was only moderate, underscoring the complexity of the task.

Two clinical features – ulcerated and flat skin surfaces – were found to be clearly associated with more aggressive tumor growth. Tumors exhibiting these characteristics were more than twice as likely to fall into one of the two higher levels of aggressiveness.

Healthcare needs should decide

The use of artificial intelligence in skin cancer care has attracted a great deal of interest in recent years, although according to Sam Polesie, so far it has had limited practical impact within healthcare. He emphasizes the importance of clearly defined application areas where research can create added value for Swedish healthcare.

We believe that one such application area could be the preoperative assessment of suspected skin cancers, where more nuanced conclusions can influence decisions. The model we’ve developed needs further refinement and testing, but the way forward is clear – AI should be integrated where it actually adds value to decision-making processes within healthcare.”

Sam Polesie, associate professor and dermatologist

Sam Polesie is an associate professor of dermatology and venereology at the University of Gothenburg and a practicing dermatologist at Sahlgrenska University Hospital. The images comprising the study data were taken within dermatological healthcare at the university hospital between 2015 and 2023.

Last year, more than 37 million people settled their behinds into the red-velvet upholstery, plastic chairs or wooden “I’ll only tolerate this because it’s the Globe” benches of a theatre. West End attendance has reportedly grown by 11% and regional audiences have increased by 4% since 2019 – pretty impressive amid a cost of living crisis and after a pandemic that had us all locked in our houses.

The increase in attendance can be chalked up to all sorts of reasons: the post-Covid return of tourists to the UK, schemes offering more reasonably priced tickets, and big films such as Wicked leaving people wondering what that Defying Gravity note sounds like live. But I’d throw another contender into the mix: the rise of AI.

For some, AI’s arrival has been exciting or, at the very least, handy – who doesn’t want to outsource life’s grunt work, or get an expert photo editor/nutritionist/therapist for nothing? For others, it feels bleak and bewildering. They’ve watched AI replace jobs, supersede human connection and infiltrate almost every area of our lives. Even worse, it’s started doing it on the sly. From AI-generated articles appearing in Wired and Business Insider (I’m real, I promise) to deepfakes of politicians going rogue, it’s becoming increasingly difficult to spot what’s real and what’s not.

That feels especially unsettling when it comes to the arts – a space where we let our emotional selves loose. It’s a sickening feeling to discover that the song that made you feel seen wasn’t written by a human with the same struggles as you. And that, no, that wasn’t real despair cracking the voice that moved you to tears.

But theatre? Sitting with other humans, watching yet more humans grapple with what it is to be human? There’s no mistaking that. Yes, the whole thing’s make-believe, but at least the artifice is out in the open. And everything else is as real as it gets, which is exactly what many of us are after.

There’s the real human connection that comes from a shared experience (no AI companions here); real points of view instead of assertions Frankensteined from every thought on the internet; real mistakes to whip Instagram’s veil of perfection from our eyes; and real variety between performances. And, of course, there are real emotions – on stage and in the audience.

That last one is especially important. In his seminal text, Poetics, Aristotle argued that feeling negative emotions while watching a tragedy not only lets us purge those emotions, but also equips us to deal with them better in our real lives. When tricky feelings can be muted with scrolling, and grief sidestepped through AI-resurrected loved ones, perhaps there’s part of us that knows that what we really need is a good old cry in a darkened room. Plus, live theatre is one of the few art forms where digital distraction just isn’t an option.

Theatre doesn’t just challenge us to feel. While AI takes the cerebral heavy lifting out of life – knowing everything so we can retain next to nothing, and telling us what to buy, eat and wear – theatre promises the opposite. At its best it holds a mirror to our greatest societal challenges and asks us what we think. What we’re going to do. A tempting proposition for anyone valiantly fighting brain rot.

But AI detractors taking solace in theatre doesn’t mean that theatre-makers have been ignoring it. AI-focused research projects are happening at Stanford University and the Royal Shakespeare Company; the National Youth Theatre has performed improv using scene prompts from Microsoft Copilot; and, next June, the Royal Ballet and Opera launches RBO/Shift, an annual festival exploring the links between opera and technology. The inaugural theme? You don’t need ChatGPT to answer that one.

AI can automate lighting and sound, generate set designs, produce live captioning and audio descriptions, and even write scripts. In the Young Vic’s 2021 production AI, a group of theatre-makers prompted GPT-3 to write one script over the course of three performances. And, in the same year, the Czech Centre in London and Prague’s Švanda theatre produced AI: When a Robot Writes a Play, a largely “autobiographical” tale.

But in true societal mirror-brandishing style, both plays interrogated the technology. In AI, audiences watched GPT-3 describe the character played by one of the actors, Waleed Akhtar, as a terrorist and typecast him as a Muslim. Guardian critic Arifa Akbar found that the robot’s autobiographical masterpiece largely consisted of it “obsessing about sex, which may not be surprising, given the prevalence of internet pornography”. Maybe theatre, then, isn’t just an escape from the perils of AI, but one of the best places to explore them in real time.

There is, of course, anxiety in the theatre community about the threats posed. In March, bodies including Equity and the Society of London Theatre co-published a manifesto aimed at protecting workers. But the technology’s potential to cut costs and streamline processes could also help the struggling sector (despite my cheery opener, theatres are facing unsustainable financial strain thanks to rising costs and shrinking investment), and many theatre-makers seem confident that AI’s role will never stretch beyond creative partner.

And, really, how could it ever hope to? Unless we reach Full Robot Takeover, no AI will ever be able to stage a play – even one that it wrote, designed and composed the music for – without those wonderful things we spoke of earlier: humans.

Theatre may just be one of the only art forms to benefit from AI without ever being truly threatened by it. Here’s hoping.

If you want to read the complete version of this newsletter please subscribe to receive The Guide in your inbox every Friday

TLDRs;

• Albania introduces Diella, an AI “minister,” to oversee public procurement amid EU pressure for anti-corruption reforms.
• Prime Minister Edi Rama says Diella will make tenders faster, more efficient, and corruption-free.
• Supporters see the AI as a step toward EU integration; critics dismiss it as unconstitutional and symbolic.
• Global examples show AI can help fight corruption but also risks bias and ineffective outcomes.

Albania has taken an unprecedented step in its long fight against corruption, introducing Diella, an artificial intelligence system tasked with overseeing public procurement.

Prime Minister Edi Rama unveiled the virtual minister as part of reforms tied to the nation’s bid for European Union membership.

Although not legally a minister under Albanian law, which requires cabinet members to be human citizens, Diella is being presented as the country’s first fully AI-powered figure in government. Her mission is clear, to bring transparency, efficiency, and accountability to one of Albania’s most corruption-prone areas.

Diella’s role in public procurement

Diella is no stranger to Albanian citizens. She first appeared as a virtual assistant on the government’s e-Albania platform, helping more than a million people navigate bureaucratic processes such as applying for official documents. Now, her responsibilities have expanded dramatically.

Rama explained that Diella’s core task will be supervising public tenders. “We want to ensure a system where public procurement is 100% free of corruption,” he said

By automating oversight and decision-making, Diella is expected to limit human interference in sensitive processes, while also making procurement faster and more transparent.

To develop this AI system, Albania is collaborating with both local and international experts, hoping to set a global precedent for AI governance.

🇦🇱 ALBANIA HIRES AI MINISTER TO FIGHT GOVERNMENT CORRUPTION

Albania has appointed Diella, an AI-powered virtual minister, to oversee public procurement and eliminate corruption from government tenders.

Prime Minister Edi Rama announced the move at his Socialist Party… https://t.co/fjDDSjsis0 pic.twitter.com/54wr4Bl5fp

— Mario Nawfal (@MarioNawfal) September 12, 2025

Mixed reactions at home and abroad

The announcement has stirred heated debate within Albania and beyond. Supporters hail the move as a chance to rebuild public trust, especially as the country faces mounting EU pressure to eliminate systemic graft.

Dr. Andi Hoxhaj of King’s College London notes that the EU has made anti-corruption reforms a central condition for accession. “There’s a lot at stake,” he said, suggesting that Diella could serve as a tool to accelerate reforms.

However, critics see the initiative as political theatre. Opposition leaders argue that branding Diella a “minister” is unconstitutional and distracts from deeper structural issues. Some worry that AI cannot fully address entrenched human networks of influence, while others raise concerns about accountability if an algorithm makes a faulty decision.

Lessons from global experiments with AI governance

Albania’s experiment comes amid a wave of governments testing artificial intelligence in public administration. Brazil’s Alice bot has reduced fraud-related financial losses by nearly 30% in procurement audits, while its Rosie bot, which monitored parliamentary expenditures, faced limitations in producing actionable evidence.

In Europe, the Digiwhist project has shown how big data can expose procurement fraud across dozens of jurisdictions. Yet, the Netherlands’ failed attempt at AI-led welfare fraud detection, widely criticized for algorithmic bias, highlights the risks of misuse.

These examples underscore both the potential and pitfalls of AI in governance. Albania now finds itself at a critical juncture: if implemented responsibly, Diella could strengthen transparency and accelerate EU integration.

Looking ahead

Prime Minister Rama acknowledges the symbolic dimension of Diella’s appointment but insists that serious intent lies beneath the theatrics. Beyond tackling procurement fraud, he believes the AI minister will put pressure on human officials to rethink outdated practices and embrace innovation.

“Ministers should take note,” Rama said with a smile. “AI could be coming for their jobs, too.”

As Albania balances hope, skepticism, and the weight of EU expectations, Diella’s debut represents both a technological leap and a political gamble. Whether she becomes a catalyst for real reform or remains a publicity stunt will depend on execution and public trust.