The Death of OpenAI's Whistleblower Makes No Sense: What Happened to Suchir Balaji?

YouTube
The legal landscape of AI training reveals a crucial divide: published works are protected by copyright, while personal data demands privacy safeguards. Recognising this distinction can shape fairer, ethical regulations for AI development worldwide.
Discover more at https://smarterarticles.co.uk/your-face-is-not-a-novel-the-moral-divide-in-ai-training?pk_campaign=rss-feed
#AIinLaw #DataProtection #CopyrightLaw #HumanInTheLoop
Your Face Is Not a Novel: The Moral Divide in AI Training

When Italy's data protection authority, the Garante per la protezione dei dati personali, slapped OpenAI with a 15 million euro fine in...

SmarterArticles

I've just finished reading the UK Government's new report today on the use of copyright works in the development of artificial intelligence (AI) systems, produced by the Dept for Science, Innovation and Technology, the Dept for Culture, Media and Sport & the Intellectual Property Office UK.

Here are the key respondent views + conclusions from the government - I'm available for on-air commentary as needed:

πŸ“ŒGov UK most concerned about AI creating "digital replicas" of someone’s voice or face - but bizarrely not bothered about copyrighted texts, images, software etc being used in AI model training datasets (the report calls it "input transparency")

πŸ“ŒThere is currently no mandate for AI developers to disclose publicly what IP they are using in their datasets, only which web crawlers they use to crawl the internet. But not what data they crawl or what it is used for

πŸ“ŒCreative industries strongly support the introduction of mandatory standards on input transparency, while tech firms have complained of the "practical and financial burdens of disclosure", particularly affecting small AI startups

πŸ“ŒAI developers & tech firms want exceptions to copyright for AI innovation. Creative industries have rejected the option to opt out of this

πŸ“ŒGov UK to work on best practices around labelling the content outputted by AI models as AI-generated, but they want to see what other nations do first

πŸ“ŒHowever, Govt UK said this despite including in this report the new legislation from California, China and South Korea that already requires AI content to be labelled as such on release

πŸ“ŒGov UK won't "intervene" in the copyright licensing market right now, because "many stakeholders" didn't want it (who tho? πŸ‘€)

πŸ“ŒGov UK to "consider" further work on barriers to enforcing IP rights

Dunno, feels like the House of Lords is right. The UK Govt is still hedging against tightening legislation, in favour of the tech giants.

But what happens when the AI models ingest everything related to all white collar jobs that require a computer? What will we do then?

Full report: https://assets.publishing.service.gov.uk/media/69ba692226909a14239612e4/CP2602959_-_Report_on_Copyright_and_Artificial_Intelligence_web.pdf

#AI #copyrightlaw #UKlaw #generativeAI #technology #technews

If you are following the #chardet storm, an even bigger one than the anti Free Software package management legislation from California, Colorado, and Illinois, this is the next place to visit:

https://github.com/chardet/chardet/issues/327

Predictably, there are a lot of people on other fora following along here and yet still missing the part where a LLM that was likely trained not only on LGPL code but also a lot of actually not #FreeSoftware code that could be scraped from various sources, was used to generate code that was then declared MIT licenced.

#CopyLeft #LLMs #VibeCoding #CopyrightLaw

No right to relicense this project Β· Issue #327 Β· chardet/chardet

Hi, I'm Mark Pilgrim. You may remember me from such classics as "Dive Into Python" and "Universal Character Encoding Detector." I am the original author of chardet. First off, I would like to thank...

GitHub
SerpApi files motion to dismiss Google's DMCA scraping lawsuit: SerpApi yesterday filed a motion to dismiss Google's DMCA lawsuit, arguing Google lacks standing as a non-copyright holder using anti-scraping tools to protect ad revenue, not creative works. https://ppc.land/serpapi-files-motion-to-dismiss-googles-dmca-scraping-lawsuit/ #SerpApi #GoogleDMCA #ScrapingLawsuit #DigitalRights #CopyrightLaw
SerpApi files motion to dismiss Google's DMCA scraping lawsuit

SerpApi yesterday filed a motion to dismiss Google's DMCA lawsuit, arguing Google lacks standing as a non-copyright holder using anti-scraping tools to protect ad revenue, not creative works.

PPC Land

How figure skating and copyright law became 'despicable' opponents, and the Canadian pair caught in middle
The figure skating world has been embroiled in an unusual kind of controversy: Copyright law has suddenly become part of an increasingly complicated artistic equation.

#figureskating #copyrightlaw
https://www.cbc.ca/sports/olympics/winter/figure-skating/milano-cortina-minions-figure-skating-chris-jones-feb8-9.7079800?cmp=rss

Seeking some light legal advice.

I'm in a community club, about 3 years old, roughly 80 members. A member of the exec board originally designed and still owns the copyright to our logo artwork, which we are well known by at this point. Some big hostility has developed among the board, so this member is resigning. He claims Title 17, Section 201(a) of US copyright law says he is the "sole owner" since he "received no compensation for the work, there was no work-for-hire agreement, and there has never been a licensing agreement". He has offered to negotiate a licensing deal, but has made it clear he intends to use it as a way to exert control over the club after his resignation or else fight our continued use of the logo.

Main question: does it seem like a simple and relatively inexpensive fight to wrest control of the logo artwork from him, legally, even if that means paying him to transfer ownership? We are registered as a 501c of some sort, I believe, in case that changes the calculus.

#copyrightLaw

Providing the publisher with the exclusive right of reproduction and distribution appears to mark the end of the author’s control over their article – but it is arguably just the beginning.

TRISTAN RADTKE on the scope and revocability of rights granted to publishers.

https://verfassungsblog.de/scope-and-revocability-of-rights-granted-to-publishers

#OpenAccess #CopyrightLaw

The source code of closed source software should be made to enter the public domain upon copyright expiry.
#software #copyright #publicdomain #opensource #closedsource #proprietarysoftware #copyrightlaw

Tweaks to IPTV Piracy Law That β€œBans VPNs” Won’t Change Its Intent or Scope

https://torrentfreak.com/govt-denies-iptv-piracy-law-bans-vpns-because-it-never-did-didnt-need-to-251218/

#copyrightlaw #Anti-Piracy #denmark #iptv #VPN

Tweaks to IPTV Piracy Law That "Bans VPNs" Won't Change Its Intent or Scope * TorrentFreak

Proposals to combat pirate IPTV led to claims that Denmark will ban VPNs. The government denies that, but will removing a term change anything?