Connect with us

ENTERTAINMENT

The Manchester Principles Put Hope V.1 on a G20 Clock

A Salford summit dated a licensed UK music model to the 2027 G20, even as 38.5 percent of July releases already showed AI and Whitehall still has no preferred.

Published

on

On 9 October 2026 a Salford summit put a UK music model on a clock that ends at Manchester’s G20. Clive Dickens, founder of Absolute Radio and now chief of The Meliora Company, named it Hope V.1 and said it should be trained, licensed and British-owned by the time world leaders arrive.

The official Manchester Principles, presented that evening at MediaCity, are milder than that pitch. They are a co-operative frame, still open on the hard machinery, issued into a market where AI already sits inside a large share of new releases.

The Wager Is a Model Called Hope

Dickens told the Creative Industries AI Summit that “the UK must own more than the playlist.” He recast GPT as Fair GPT: fair governed, fair powered and fair trained. The proposed model would pull on British knowledge that still sits off the open web, pay the people whose work trains it, and be ready when Manchester hosts the G20 on 27 and 28 November 2027.

People say we’re too late. It’s not too late: we are in the first five minutes of the next industrial revolution. We can train our own model.

Clive Dickens, founder and CEO, The Meliora Company, at Beyond The Music

He called AI “a basic human right, in the same way that water, shelter and power are,” and said about 2 billion people now use it every week in a form that “didn’t really exist 50 months ago.” He put the human population at 8.2 billion. The Fair GPT brief matches that scale with three demands: rules before release, renewable power for the data centres, and no scraping without provenance, consent and pay.

Hope V.1 is still a speech, not a lab with a budget line. Dickens said so himself, giving it a hypothetical name and a date. On 10 November 2026, The Meliora Company said it had backed five more generative AI startups across Australia, the UK, Ireland and France, taking its venture book to 17 firms, with at least eight of them in generative AI. The person dating a licensed British model is also writing cheques into the same trade.

Ian Murray, the minister for creative industries, opened the day by saying “we cannot ask our creators to choose between embracing innovation and protecting their rights.” Bev Craig, Mayor of Greater Manchester, asked how economic and technical change could still “protect creativity and our very soul.” Prime Minister Andy Burnham, a Beyond The Music founder, convened the summit as part of a festival that ran from 6 to 9 October 2026. The bet now sits in public, with a city and a calendar attached.

Five Rules the Room Could Agree On

Rose Marley, chair of Beyond The Music, and Cliff Fluet, the summit chair and a media lawyer, presented the Principles at the close. The site that now holds them is clear about what they are not. They are not a statute, and no commercial party with a material interest is meant to hold the text.

The argument began in a white paper, The Enclosure of the Human Mind, then moved through a co-operative tradition that Marley and Oli Wilson of Beyond The Music took to the mayor’s office. An August 2026 working draft ran to 29 clauses under four pillars: licensing, provenance, attribution and enforcement. On the day, that draft was boiled into a frame.

WHAT THE ROOM KEPT REPEATING

  • Human first: Human creativity comes first, ahead of the model that trained on it.
  • Ask before use: Consent means asking before someone’s work, voice or likeness is used.
  • Pay the source: Fair pay for the work that AI is built on, not a courtesy credit.
  • Show the trail: Transparency about what has been used to train and make things.
  • Smallest player: Enforcement that works for the smallest rights-holder, not only the largest group.

Those five lines sat under five co-operation rules from the room: work together in good faith, share value if the work makes the money, respect creators, invest in the next generation, and keep the table open to people who do not agree on every detail. The site says the other 20 percent is where the interesting work remains, and a working group, capped at 1,000 places checked by a person, is meant to take that on, with a Manchester session planned before Christmas 2026. The University of Southampton’s Web Science Institute has opened a collaboration on the mechanisms behind the frame.

What Share of New Music Already Uses AI?

SubmitHub’s SH Labs detector scanned more than a million releases from July 2026 and classified 38.5 percent as showing AI involvement: 23.2 percent fully generated, 15.3 percent hybrid, with AI audio later edited by a person. Among artists whose tracks were flagged, 31 percent said they had not used AI at all. Jason Grishkoff, SubmitHub’s founder, said the path forward is disclosure, so listeners can decide what they want to hear.

THE FLOOD IN TWO DATASETS

Measure Figure Who counted it
July 2026 releases with some AI 38.5% SubmitHub SH Labs, 1 million-plus tracks
Fully AI-generated share 23.2% SubmitHub SH Labs
Hybrid AI-plus-human share 15.3% SubmitHub SH Labs
Flagged artists who denied using AI 31% SubmitHub follow-up
Listeners who failed a blind test 97% Deezer and Ipsos, 9,000 people

A Deezer and Ipsos survey fielded 6 to 10 October 2025 asked 9,000 adults in eight countries, including Britain, to pick the human track from a set of three. 97 percent failed a three-track blind test. Of those respondents, 71 percent were surprised and 52 percent felt uncomfortable. Some 80 percent wanted fully AI-generated music clearly labelled, 65 percent said copyrighted material should not be used to train music models, and 70 percent thought fully AI-generated music threatens the livelihood of working musicians. Deezer also said it was receiving about 50,000 fully AI-generated tracks a day, 34 percent of daily deliveries, though those tracks were about 0.5 percent of streams, with up to 70 percent of those streams judged fraudulent.

That is the catalogue Hope V.1 would have to enter, not a future one. Streamers are already burying AI-made work rather than leaving it in the main mix; Spotify hid 50 million AI-made tracks from recommendations without deleting them. Ian Hogarth, co-founder of Songkick and chair of the AI Security Institute, said on 10 November 2026 that “what we are doing doesn’t really seem to be working to protect UK creatives or enable UK AI companies.” The current path is serving neither camp well, which is why a dated British model is being sold as more than a slogan.

The March U-Turn Left a Blank Page

Murray cited the threat number because it is now a sector fact. UK Music’s This Is Music survey, published 27 November 2025, found that 66 percent of creators see a career threat from AI. Only 13 percent disagreed, and 21 percent sat in the middle. Over 90 percent wanted consent, pay, labelling, and protection for voice and image. The same report put the UK music industry’s 2024 economic contribution at a record £8 billion, with exports at £4.8 billion, even as the growth rate slowed by more than half on 2023.

CISAC, the global network of authors’ societies, and PMP Strategy warned in December 2024 that 24 percent of music revenues at risk by 2028, a cumulative €10 billion over five years and €4 billion in 2028 alone. They put the market for AI-made music and audiovisual content on a path from about €3 billion to €64 billion by 2028, with music AI services rising from €0.1 billion in 2023 to €4 billion. By 2028, they said, AI-made music could account for about 20 percent of streaming-platform revenues and about 60 percent of music-library revenues. Creators lose twice in that model: unpaid training, then substitution on the same shelves.

HOW THE COPYRIGHT FIGHT GOT TO SALFORD

  1. May and June 2025: Baroness Beeban Kidron’s transparency amendments to the Data (Use and Access) Bill pass the Lords, then fall. The Commons rejects them by 304 votes to 189, and the Act receives Royal Assent without a duty on AI firms to disclose training data.
  2. December 2025: A government consultation on copyright and AI draws about 11,500 responses. Around 3 percent back a broad text-and-data-mining exception with an opt-out.
  3. 18 March 2026: Technology Secretary Liz Kendall publishes the required progress report and dropped the opt-out exception as preferred. The government says it no longer has a preferred option and will not add new copyright exceptions for now. Existing law still says copyright material cannot be used for training without permission.
  4. August 2026: The Manchester Principles working draft, 29 clauses, is published as a starting point, not a finished text.
  5. 9 October 2026: The Principles are presented at MediaCity Salford. The open questions stay open.
  6. 27 and 28 November 2027: The G20 leaders’ summit is due at Manchester Central Convention Complex, with up to 10,000 attendees, and Burnham has said AI will sit at the heart of the UK presidency.

Tom Kiehl, chief executive of UK Music, called the March shift a win and then asked ministers to rule the opt-out out for the rest of the term. The Ivors Academy called it a first step that avoided the worst outcome. The blank page is the point of Salford: the industry wrote a frame because Whitehall would not write a rule.

A Birth Certificate That Travels in the File

Andrew Melchior, chief technical officer for Massive Attack and chief of Genotone, told a panel the sector needs “clarity, standards and standardised tools that avoid bespoke walled gardens.” His firm embeds an inaudible mark in the audio at mastering, binds it to a verified person, and treats that mark as a birth certificate that survives compression and stripped metadata. He has said the goal is to “embed artist rights at the point of creation of a piece of music that travels through the system with it.”

Proof-of-concept tests put the mark inside Massive Attack’s Teardrop and Bicep’s Glue, with the artists unable to hear it, then inside John Cage’s four minutes 33 seconds of silence. If it holds in silence, Melchior argues, it holds. He is not selling an AI detector. Detectors guess and can false-flag a human mix; his system verifies a signed declaration by a named person. Sir Robin Jacob, a former appeal judge in intellectual property, has joined Genotone’s advisory board. The prototype still needs a standard the rest of the industry will accept, which is the public-infrastructure problem the Principles named and did not solve.

WHAT FAIR GPT WOULD REQUIRE

  • Fair governed: Pre-approval before a model ships, on the same logic as a medicine, plus reporting and a pact against harmful systems.
  • Fair powered: Renewable energy for the data centres that train and run the models.
  • Fair trained: No scraping without provenance, consent and pay, so the people in the training set can earn from the output.

Kiehl, on the same panel, called mass-scraping, consent, deepfakes and AI slop “huge” and “deeply personal to artists.” “This is about how your work can be manipulated, how your work can be changed, how your work can be attributed in a different way,” he said. The missing market, if labs will not negotiate song by song, looks like a collecting society for training: one licence, usage data, a split. That is the job the 1,000-person working group has been asked to start, without a statute behind it.

Kidron Warns That Meaning Goes With Control

Kidron, the crossbench peer and former film director who led the 2025 Lords fight, with backing from Elton John, Paul McCartney and Dua Lipa, told the Salford room that the loss is not only a royalty line. When someone writes, films, records or paints, “someone made a choice, someone wanted us to hear or feel something we otherwise might not have heard or felt.” That intention, she said, is “stripped of value in this race to control artificial intelligence.”

The threat is existential. Lose copyright and you lose the means of earning a living, lose control of your work and you lose control of its meaning.

Baroness Beeban Kidron, crossbench peer, at Beyond The Music

She wanted ownership and consent, transparency and traceability, fair licence and new markets at the centre of any future deal. “The danger is that if we don’t have an agreement about what we want, then there are people out there who definitely know what they don’t want us to have.” A sovereign model, in her telling, is less a patriotic gadget than a way for Britain to set and enforce its own terms instead of taking them from labs and their home governments.

She described a teenage boy who told her “the future is disappearing and he could no longer see himself in it.” “Hold the next generation in mind,” she said, “because already the prospect of human replacement is making children lose sight of their future.” The tech lobby’s line, that the tools move too fast to govern, is a choice. “The future hasn’t disappeared, it’s being made right now. And as every creator knows, what is made can be made differently.”

Manchester Hosts the Clock Next November

Fluet closed the day by saying the UK music business wants “creative sovereignty” because “our culture is a shared common wealth and it’s quietly wearing away.” Consent, fair pay, transparency and enforcement that works, or the catalogue of the past, present and future gets stripped. “Whatever we build has to feed back into the society it’s built from, rather than taking from it.”

Just over a year separates the Salford room from the G20 at Manchester Central. The working group is still forming. Hope V.1 has a name and a date, not a training run. Kendall’s March report left the legal option blank. SubmitHub’s July scan and Deezer’s daily intake say the catalogue is not waiting. The Principles hold the picture. They do not paint it, and the G20 will arrive on 27 and 28 November 2027 whether the model does or not.

Harry is the editor of THE iBULLETIN, an independent publication he owns and runs. He has been in journalism for ten years, first reporting and later editing, and much of what the site covers now begins in its inbox. Reader mail is read in full, every message of it. A tip is treated as a lead to be verified, not a story to be printed, and a challenge to a published fact is checked against the original filing, statement or transcript within the day, with the article corrected under a public policy if the reader is right. Questions that several readers ask become articles. That exchange feeds coverage of news, business and technology, of science and sports, and of entertainment, lifestyle, travel, auto and gaming, written for readers spread across many countries rather than one. Harry works from primary sources and checks each number himself before publication, and he would rather run a shorter story than an unconfirmed one. The address for all of it, tips, corrections and questions alike, is support@theibulletin.com.

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending