#gpai — Public Fediverse posts

OpenAI publishes its EU AI Act compliance blueprint as GPAI enforcement looms

If this matters to you, share it.

https://1ban.news/openai-gpai-code-compliance-2026/

#1ban #openai #gpai #code #compliance #tech

OpenAI has endorsed both the EU's GPAI Code of Practice and Transparency Code, and detailed the internal frameworks it says sit behind those commitments. https://www.artificialintelligence-news.com/news/openai-aligns-safety-practices-with-eu-ai-act-gpai-code/ #openai #aiact #governance #enterpriseai #eu #gpai #europe #ai #tech

Druhého srpna začnou v EU platit povinnosti pro poskytovatele i provozovatele high-risk AI systémů: posouzení shody, technická dokumentace a quality management na straně providerů, uchovávání logů a dohled nad provozem na straně deployerů. Samostatně vstupují v platnost transparentní pravidla pro chatboty, generativní AI a deepfaky, a ta se týkají všech, nejen high-risk systémů. Kdo nasazuje AI v recruitmentu, credit scoringu nebo HR hodnocení, je v zóně. Čekání na odklad přes Digital Omnibus je sázka na legislativní proces, který ještě neskončil. A kdo si myslí, že se ho to netýká, protože „jen používá ChatGPT“ v use casu z Annexu III, pravděpodobně špatně přečetl nařízení.

EU AI Act и санкции: двойной регуляторный барьер для российских IT-компаний

13 марта 2026 года Совет ЕС согласовал позицию по Digital Omnibus on AI — пакету упрощения правил EU AI Act. Через пять дней, 18 марта, в России опубликован проект федерального закона «Об основах государственного регулирования применения технологий искусственного интеллекта». Два регуляторных режима с фундаментально разной философией входят в решающую фазу одновременно. Для IT-компаний из РФ, экспортирующих SaaS- и B2B-продукты на европейский рынок, ситуация осложняется дополнительным фактором — 19 пакетами санкций ЕС против России. С декабря 2023 года запрещена поставка корпоративного ПО в Россию, а с ноября 2025 — предоставление AI-сервисов. EU AI Act добавляет к этому санкционному каркасу ещё один слой требований: экстерриториальное регулирование, обязательную классификацию рисков и штрафы до 35 млн евро. Статья разбирает механику EU AI Act, его пересечение с санкционным режимом и российским законодательством, а также практические последствия для компаний, работающих на стыке двух юрисдикций.

https://habr.com/ru/articles/1015812/

#EU_AI_Act #sanctions #GPAI #AI_regulation #compliance #SaaS #IT #Omnibus_VII

The European Parliament adopted Resolution P10_TA(2026)0066 on copyright and generative AI on 10 March 2026.

Three critical issues — but not resolved:

▸ The structural failure of the DSM Directive's opt-out (Art. 4)
▸ Transparency obligations extended to inference and RAG
▸ A regulatory architecture built on voluntary self-discipline

The real test will be the Commission's response.

https://www.nicfab.eu/en/posts/copyright-genai-ep-resol/

#Copyright #GenerativeAI #AIAct #DSMDirective #GPAI #TDM #AI #artificialintelligence

Scraping artists, creators, and right holders with...

No transparency.
No traceability.
No enforceability.
No Accountability.
= No functioning copyright market.

#AIAct #CreativeIndustries #Copyright #Audiovisual #Publishing #GPAI #Netopia

Новые правила для GPAI и «каскад обязанностей»: как небольшой команде превратить риски EU AI Act в преимущество

Недавно опубликованные Еврокомиссией требования к моделям общего назначения (GPAI) по Закону ЕС об ИИ (EU AI Act) задают правила игры для всех, кто работает с генеративным ИИ. Даже если вы — небольшая команда, которая интегрирует в свой продукт API от OpenAI, использует Llama или любую другую foundation model, эти новые нормы касаются вас напрямую как Деплоер. Важно понимать: речь не о том, чтобы отвечать за ошибки Google или Anthropic. Речь о том, что у Деплоера появляется своя, отдельная зона ответственности. Но для небольшой и гибкой команды это не только новые риски, но и уникальная возможность. В этой статье мы посмотрим на проблему с точки зрения небольшой компании разрабатывающей свой ИИ продукт и разберем, что входит в эту зону ответственности и, самое главное, как использовать новые обязанности «больших компаний» как рычаг для построения более надежного и конкурентоспособного продукта .

https://habr.com/ru/articles/958010/

#GPAI #комплаинс #ai #ai_governance

https://technologiesinternetz.blogspot.com/2025/10/global-partnership-on-artificial.html

Global Partnership on Artificial Intelligence (GPAI): The Catalyst for Revolutionary Change

#GlobalPartnershiponArtificialIntelligence #GPAI #Artificialintelligence #AI

Stellungnahme des Deutschen #AnwaltVerein (#DAV) zum Initiativberichtsentwurf #Urheberrecht und generative #KI

Der DAV sieht Nachbesserungsbedarf in Bezug auf die Regulierung von urheberrechtlichen Aspekten bei generativer künstlicher Intelligenz.

Download als PDF: https://anwaltverein.de/newsroom/sn-58-25-dav-stellungnahme-zum-initiativberichtsentwurf-urheberrecht-u-generative-ki

#genAI #UrhR #AI #GPAI

https://technologiesinternetz.blogspot.com/2025/09/unpacking-gpai-your-essential-guide-to.html?m=1

Unpacking GPAI: Your Essential Guide to the Global Partnership on Artificial Intelligence

#GlobalPartnershiponArtificialIntelligence #gpai #artificialintelligence #ai

https://technologiesinternetz.blogspot.com/2025/09/unpacking-gpai-your-essential-guide-to.html?m=1

Unpacking GPAI: Your Essential Guide to the Global Partnership on Artificial Intelligence

#GlobalPartnershiponArtificialIntelligence #gpai #artificialintelligence #ai

Our Equity & Data programme director Laura Lazaro Cabrera talked about the #AIAct & the Code of Conduct for #GPAI models during the latest episode of @euronews's EU Decoded, emphasising the need of a strong safety framework that promotes fundamental rights as well as people-centred #AI systems.

👇🏻 Watch the video & read the article here: https://www.euronews.com/my-europe/2025/08/19/will-regulation-make-the-eu-the-most-trusted-power-in-ai

Aug 1, 2025: Just 24h before AI Act's second phase, @EUCommission declares the General-Purpose AI Code "adequate."
Key findings:
- 12 commitments reshape AI governance
- "Similarly safe models" add flexibility
- Independent evaluators create new markets
- Continuous monitoring with emergency protocols

The future of European AI is being implemented.
📖 Full analysis: https://www.nicfab.eu/en/posts/ai-act-eucom-approves-code-of-practice/
It follows my previous post: https://www.nicfab.eu/it/posts/ai-act-second-apply/

#AIAct #GPAI #EuropeanCommission #AIGovernance #AI

Datenschutzbehörde Hamburg: "#KI-Verordnung nimmt #Anbieter in die Pflicht"

-> "Die neuen Pflichten gelten für alle #GPAI-Modelle, die nach dem 2. August 2025 auf den europäischen Markt kommen"

https://datenschutz-hamburg.de/news/ki-verordnung-nimmt-anbieter-in-die-pflicht

🚀 The last #AIBulletin before summer break is here!

This edition features:

📢 #GPAI Code of Practice

📃 #AI Act’s guidelines & transparency template for GPAI models

🗣️ Intensified debate over the AI Act’s delay & simplification

🙋🏻 AI Office seeking new advisory forum’s members

👇 Read more & sign up:
https://cdt.org/insights/cdt-europes-ai-bulletin-july-2025/

Eine Woche vor Inkrafttreten der neuen Vorschriften für künstliche allgemeine Intelligenz (GPAI), wird klar, wo die Unternehmen stehen, wenn es darum geht, den freiwilligen #Verhaltenskodex der #EU für #GPAI zu unterzeichnen.
https://de.euronews.com/my-europe/2025/07/24/meta-weigert-sich-ki-kodex-der-eu-wird-nicht-unterzeichnet

📢 Yesterday, the European Commission published the long-awaited final draft of the Code of Practice for general-purpose AI (#GPAI) models.

We welcomed it as an improvement in the taxonomy, now requiring providers to identify systemic risks for assessment based on broader risk categories.

However, it continues to deprioritise many fundamental rights concerns by leaving it to providers to identify and assess these risks.

👇🏻 Read our full statement: https://cdt.org/insights/cdt-europes-reaction-to-the-final-code-of-practice-on-general-purpose-ai-draft/

📢 The last EU Tech Policy Brief before the summer break is here!

This issue covers:

🚨 CDT Europe’s take on the CoP on #GPAI

👁️ Letter on #spyware abuses in the EU

💬 Feedback on online protection of minors guidelines

⚖️ Letter against the reopening of the #AIAct

👇 More about EU tech policy:
https://cdt.org/insights/eu-tech-policy-brief-july-2025/

How #US Firms are weakening the #EU #AI #Codeofpractice: By pressuring the European Commission to prioritize a few US firms over 1,000 stakeholders, these companies put the entire process at risk and lose credibility as actors of public interest #AIAct #GAFAM # OpenAI #GPAI #AICodeofPractice #Code
https://www.techpolicy.press/how-us-firms-are-weakening-the-eu-ai-code-of-practice/

On 19 May, CDT Europe organised a CSOs workshop on the #AI Act implementation.

The workshop focused on concerns with and lessons learned from the Code of Practice on #GPAI models, national enforcement structures, & civil society action in light of the Commission’s simplification agenda.

👇🏻 Read more
https://cdt.org/insights/cdt-europe-hosts-a-second-civil-society-workshop-to-discuss-the-ai-acts-implementation/

🧠 AI hallucinations are not just technical flaws but a governance risk.

Lloyd’s of London now offers insurance against AI-generated errors, including hallucinations, via a policy developed by Armilla AI.

As AI becomes central to decision-making, legal and ethical safeguards are essential.

🌐 Source: https://www.business-standard.com/industry/news/insurance-chatbot-armilla-lloyds-ai-hallucination-artificial-intelligence-125051200951_1.html

#AI #AIHallucinations #GPAI #ResponsibleAI #artificialintelligence #LLMs

🇪🇺 Safeguarding culture in the age of AI

Italy, Portugal, and Spain call on the EU Council to protect the cultural and creative sectors in the development of GPAI.

🎯 Priorities:
• Copyright
• Transparency on training data
• Legal certainty (DSM opt-outs)
• Alignment with EU values

The AI Act & upcoming Code of Practice must not leave creators behind.

🌐 Source: https://data.consilium.europa.eu/doc/document/ST-8188-2025-INIT/en/pdf

#DigitalCulture #AIAct #GPAI #Copyright #AI #artificialintelligence

"Simplification should not be a guise for deregulation.

By mistaking transparency and openness for an obstacle, not a driver, of innovation, it would shoot itself in the foot. The new transparency rules for AI and data under the EU’s AI Act may become one of the first casualties of this new impetus to roll back some of the recently adopted requirements for the providers of so-called general-purpose AI (GPAI) models.

Under the EU’s AI Act, developers of GPAI models — that is, very large AI models such as OpenAI’s GPT or Google’s Gemini models — will soon have to present a “sufficiently detailed” public summary of the data they used to train the models.

This summary could be a light-touch way to drastically advance transparency around the use of one of AI’s most precious inputs, data at little additional cost to developers.

But if the EU’s AI Office gives in to industry pressure to water down the level of detail, this summary will turn into a performative checkbox exercise that ultimately offers little value to anyone. This would be misguided and short-sighted."

https://euobserver.com/Digital/ara7bbd63b?utm_source=EUobserver+Newsletter&utm_campaign=9f48e6cc56-EMAIL_CAMPAIGN_2025_04_17_05_23&utm_medium=email&utm_term=0_8f318ca8da-9f48e6cc56-114913537

#EU #AI #AIAct #GPAI #AITransparency #Deregulation

"The substance of the draft is equally concerning. While its purpose is to help providers comply with existing obligations, the current draft goes beyond mere clarification – introducing new requirements not envisioned in the AI Act.

One example is the proposed role of “external evaluators” before releasing GPAI models with systemic risks, which is not provided for in the AI Act. The draft mandates providers to obtain external systemic risk assessments, including model evaluations, before placing their models on the market (commitment II.11). However, the AI Act itself (Art. 55(1)(a) and recital 114) does not impose this requirement – it only calls for adversarial testing of model evaluations, not independent external risk assessments.

Another example concerns copyright: measure I.2.4. of the draft requires GPAI model developers to make reasonable efforts to determine whether protected content was collected by a robots.txt-compliant crawler – an obligation not imposed by the AI Act either. Additionally, measure I.2.5. mandates that GPAI model providers take reasonable steps to mitigate the risk of downstream AI systems repeatedly generating copyright-infringing content and to prohibit such uses in their terms and conditions. However, these requirements are not found in the AI Act or the Copyright Directive 2019/790, which addresses only primary liability (i.e., the responsibility of GPAI model providers) and does not extend to secondary liability arising from text and data mining.

Again, the issue is not whether these requirements are reasonable, but that the Code’s sole purpose is to clarify the obligations of the AI Act, not to redefine them. Therefore, the Code must not be used as a Trojan horse to reshape the AI Act according to political preferences – bypassing democratic procedures."

https://verfassungsblog.de/when-guidance-becomes-overreach-gpai-codeofpractice-aiact/

#EU #AI #AIAct #CodeOfPractice #GPAI

🧠 #AIhallucinations: How Can #Privacy Be Protected?

⏯️ Professor Christakis explains how focusing on #GPAI outputs – rather than internal workings – can help protect #DataProtection while fostering #AI innovation.

#AIact #EuropeanAIroundtable

https://www.youtube.com/watch?v=We-waT7FOWc

"WITNESS expresses concern regarding the changes in the EU AI Act’s third draft of the General-Purpose AI Code of Practice (CoP). These modifications aim to lessen the burden of applying the EU AI Act on General-Purpose AI (GPAI) model providers. This directly impacts transparency, risk assessments and copyright, and undermines fundamental rights. We also believe these changes negatively impact the information environment and the ability to detect the use of AI and ensure content authenticity. They prioritise reduced implementation over obligations towards more robust rights protections.
(...)
While one of the goals of the document is to provide a detailed framework for AI providers to ensure their models are safe, secure, and compliant with relevant regulations, we have concerns with a reduction of emphasis on some obligations now present in the latest version of the CoP, as well as a possible reduction in the scope of the systemic risks to be mitigated by GPAI providers."

https://blog.witness.org/2025/03/eu-ai-act-ensuring-rights-disclosure/

#EU #AI #AIAct #FundamentalRights #GPAI #GenerativeAI

I'm forcing my way through ver 3 #GPAI #CoP – which actually has improved, but the #AGI stuff is still painful. I saw @samasama.bsky.social at #AIActionSummit say the definition of AGI was meaningless at this stage, but I've heard someone promised Trump AGI would happen in this term. Do we know who?