Elzatian AI animator · AI content creator · AI trainer

AI news · creativity

Why the Pope says AI art isn't art

Because he is arguing about where a picture comes from rather than how good it looks. Meeting artists and curators in the Apostolic Palace on 2 October 2026, Pope Leo XIV said the Church needs them, 'especially true in this age of artificial intelligence, in which we urgently need to distinguish human art from what is produced by machines', and that the difference is 'ontological — even before an aesthetic one — between art and what a machine can generate through statistical calculation based on millions of images created by others'. The same week, the tools creators are actually being handed pointed elsewhere: Meta's Edits Assistant, rolling out to creators in the United States from 30 September 2026, reads an account's own views, retention, likes and shares and suggests what to try next, while leaving the creative decisions to the person, by design. Neither the praise nor the worry is new; what is new is that both are now public at once, and that nobody has measured which one audiences prefer. Singapore has not measured it either, and says a study is expected in 2027.

7 min read 6 sections 3 October 2026 Written by Elza

Summary

On 2 October 2026 Pope Leo XIV met artists and curators at the Vatican and drew the line where he has drawn it before: between art a person makes and what a machine generates from millions of images made by others. 'There is an ontological difference, even before an aesthetic one,' he said, and algorithms lack 'the spark of the human'. In the same week the tools creators are being handed went another way. Meta began rolling its Edits Assistant out to creators in the United States from 30 September 2026, reading a creator's own Reels numbers and explaining what worked, and YouTube said its own conversational editing tool arrives next year. Our view is that the Pope's claim is about where a picture comes from, while the tools arriving are about measurement, and that Singapore has said plainly it is still studying the question.

What the Pope actually said

The Pope was speaking in the Apostolic Palace at a meeting called The Art of Possibility: Weaving Together, organised by the Dicastery for Culture and Education, with artists and curators in the room. He opened by telling them the Church needs them, 'especially true in this age of artificial intelligence, in which we urgently need to distinguish human art from what is produced by machines'. The distinction he drew is one he has drawn all year: there is 'an ontological difference — even before an aesthetic one — between art and what a machine can generate through statistical calculation based on millions of images created by others', and 'algorithms lack the spark of the human', whereas 'authentic culture and art preserve this spark'. That phrase is his own, taken from his encyclical on AI, Magnifica Humanitas, published on 15 May 2026. He added a comparison aimed at the difference between answering and making: in an age of search engines and their endless, instantaneous answers, he said, art gives us something other than what we already want. The same argument went out from his X account that morning, and the interest has not faded — the New York Times reported on 29 September 2026 that Anthropic had pressed the Vatican to reconsider its position on non-human consciousness.

The claim is about origin, not finish

It is worth being exact, because the sentence is easy to over-read. The Pope is not saying machine images look bad, or that they are derivative in a way a court could test. He is saying the difference is not aesthetic at all: before you judge the picture, note that nobody behind it lived the thing it shows. That is a claim about authorship, offered as principle rather than finding, and the part with a price attached is the corollary. If the raw material really is 'millions of images created by others', then whoever supplied those images is inside the argument whether or not they agreed to be.

The tools arriving analyse rather than create

The counterweight came from Meta two days earlier. Instagram began rolling an assistant out inside its Edits video app, the CapCut rival, to creators in the United States from 30 September 2026. Edits Assistant reads an account's own follows, views, video retention, likes and shares, combines them with comments, platform trends and audience signals, and answers questions about them: why one Reel held people and another lost them early, or what to try next. Its head, Brett Westervelt, described it as pattern-spotting 'that is hard to see from the metrics alone', and reported what the creators he tested with kept asking for: 'creators want a tool that handles the analysis, not one that does the creative work for them. Edits assistant does the digging for you. The creative calls are still yours.' It suggests concepts, hooks, captions, scripts and trending audio, free inside the app with a daily limit on prompts, with more access through Meta One. YouTube said the week before that it is building a conversational editing tool of its own. The two moves point opposite ways: the Pope is arguing about who makes the picture, while the platform that hosts most of them is automating the judgement about what to make next.

Where Singapore's own answer stands

There is no local number to cite on this question, and it is worth saying so plainly. Answering parliamentary questions on 6 August 2026, the Ministry of Culture, Community and Youth said it had been engaging the arts community and found practitioners split: some raised concerns about the impact on livelihoods and about legal and ethical issues, while others said AI can serve as a collaborative tool for ideation, experimentation and new forms of expression. It has not commissioned its own study; research run with the Institute of Policy Studies Policy Lab is expected to publish findings in the first half of 2027. The National Arts Council's practical work here is legal rather than aesthetic: intellectual property resources for self-employed arts workers, a partnership with Pro Bono SG widening access to legal advice including on AI issues, and an Arts x Tech Lab for practitioners who want to experiment. The Ministry's wording is a policy position rather than an assessment of the art — technology should complement, not replace, 'the originality, lived experience, and creative judgement that artists bring to their work'.

Our take

We are sympathetic to the Pope's line and suspicious of how easily it gets repeated. That a machine works from other people's images is true and worth saying; that this makes the result not art is a judgement, not a finding, and the same sentence was aimed at photography when it was new, and at the synthesiser later. The testable question is narrower: can audiences tell, do they care, and does the person who made the thing get paid. On that last point he is on firmer ground than he is on aesthetics, because the training corpus is a consent problem wearing the clothes of a philosophy problem.

We are more interested in what Meta built, because it is the opposite bet. Edits Assistant does not make the film; it reads your retention curve and tells you why the last one lost people early. That is the part of the work most creators dislike, and handing it over does not touch authorship at all. The honest risk is the obvious one: a platform advising creators to make more of what the platform already rewards tends to produce more of the same, and 'trending audio' is not a point of view.

For a studio here, neither side of the argument pays the bills. We would use the analysis tools without pretending they are creative partners, and keep a paper trail on where our own images came from — which, as the arts council's resources and the legal clinics it funds will tell you, is becoming a business requirement rather than a matter of taste.

Where this is taught

AI Film Lab: Cinematic Storytelling — 2 Days, taught live in a small cohort. Ask on WhatsApp (+65 9188 6948) for the next dates.

See the course Hand me a brief

Keep reading

The other articles

Choosing a course

Which AI course in Singapore is worth it?

We judge an AI course in Singapore by what you leave with, not by the syllabus. The ones worth paying for end with something you made on your own material and a method you can run again without the teacher; the rest end with a certificate and a folder of clips you cannot explain.

Read more

Getting started

What to learn first in AI video

Learn the decisions before the tools. The order we use is: write the shot list, lock the character and the look, approve a still, then animate, then finish the sound — because each stage makes the next one cheaper.

Read more

Funding

Don't spend SkillsFuture Credit yet

SkillsFuture Credit can be used for courses listed on the MySkillsFuture portal, and eligibility is decided per course rather than per subject — so a course being about AI tells you nothing about whether it is claimable. Before you commit, confirm three things: that this specific course is listed as eligible, that the subsidy in the quote applies to you, and that you know what you leave with.

Read more

Commercial work

What should a small business film first?

Three videos earn their keep before anything else: a short product spot that shows the thing working, a founder or team story that explains why anyone should care, and vertical cut-downs of both for social. Brand films, event recaps and animations are later purchases — worth making once those three are already doing their job.

Read more

Short-form

AI videos that don't look fake

AI short-form reads as fake for three reasons, and picture quality is not one of them: the first second is a logo instead of a sentence, the face changes between cuts, and the sound does not belong to the room. Fix those three and an AI-made video can sit in a feed without being spotted — which is the only thing that makes it worth making.

Read more

AI news · creativity

Why AI music is flooding streaming

Because supply became almost free and attention did not. Deezer reported on 21 July 2026 that fully AI-generated tracks reached a monthly average of about 90,000 a day in June, more than half of everything uploaded to it at peak, and only 1% to 3% of its actual listening; in 2025 it detected and tagged 13.4 million such tracks.

Read more

The method behind these pieces is written out in the guides — the six stages, the four rules behind every prompt, and how a character survives a whole film.