The General-Purpose AI Code of Practice seeks to clarify legal obligations under the act for providers of general-purpose AI models such as ChatGPT, especially those posing systemic risks like ones that help fraudsters develop chemical and biological weapons.

The code’s publication “marks an important step in making the most advanced AI models available in Europe not only innovative but also safe and transparent,” Henna Virkkunen, executive vice president for tech sovereignty, security and democracy for the commission, which is the EU’s executive arm, said in a statement.

The code was developed by 13 independent experts after hearing from 1,000 stakeholders, which included AI developers, industry organizations, academics, civil society organizations and representatives of EU member states, according to a Thursday (July 10) press release. Observers from global public agencies also participated.

The EU AI Act, which was approved in 2024, is the first comprehensive legal framework governing AI. It aims to ensure that AI systems used in the EU are safe and transparent, as well as respectful of fundamental human rights.

The act classifies AI applications into risk categories — unacceptable, high, limited and minimal — and imposes obligations accordingly. Any AI company whose services are used by EU residents must comply with the act. Fines can go up to 7% of global annual revenue.

The code is voluntary, but AI model companies who sign on will benefit from lower administrative burdens and greater legal certainty, according to the commission. The next step is for the EU’s 27 member states and the commission to endorse it.

Read also: European Commission Says It Won’t Delay Implementation of AI Act

Inside the Code of Practice

The code is structured into three core chapters: Transparency; Copyright; and Safety and Security.

The Transparency chapter includes a model documentation form, described by the commission as “a user-friendly” tool to help companies demonstrate compliance with transparency requirements.

The Copyright chapter offers “practical solutions to meet the AI Act’s obligation to put in place a policy to comply with EU copyright law.”

The Safety and Security chapter, aimed at the most advanced systems with systemic risk, outlines “concrete state-of-the-art practices for managing systemic risks.”

The drafting process began with a plenary session in September 2024 and proceeded through multiple working group meetings, virtual drafting rounds and provider workshops.

The code takes effect Aug. 2, but the commission’s AI Office will enforce the rules on new AI models after one year and on existing models after two years.

A spokesperson for OpenAI told The Wall Street Journal that the company is reviewing the code to decide whether to sign it. A Google spokesperson said the company would also review the code.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

Read more:

Severe weather events, such as heavy rainfall, are on the rise worldwide. Reliable assessments of these events can save lives and protect property. Researchers at the Karlsruhe Institute of Technology (KIT) have developed a new method that uses artificial intelligence (AI) to convert low-resolution global weather data into high-resolution precipitation maps. The method is fast, efficient, and independent of location. Their findings have been published in npj Climate and Atmospheric Science.

“Heavy rainfall and flooding are much more common in many regions of the world than they were just a few decades ago,” said Dr. Christian Chwala, an expert on hydrometeorology and machine learning at the Institute of Meteorology and Climate Research (IMK-IFU), KIT’s Campus Alpin in the German town of Garmisch-Partenkirchen. “But until now the data needed for reliable regional assessments of such extreme events was missing for many locations.”

His research team addresses this problem with a new AI that can generate precise global precipitation maps from low-resolution information. The result is a unique tool for the analysis and assessment of extreme weather, even for regions with poor data coverage, such as the Global South.

For their method, the researchers use historical data from weather models that describe global precipitation at hourly intervals with a spatial resolution of about 24 kilometers. Not only was their generative AI model (spateGEN-ERA5) trained with this data, it also learned (from high-resolution weather radar measurements made in Germany) how precipitation patterns and extreme events correlate at different scales, from coarse to fine.

“Our AI model doesn’t merely create a more sharply focused version of the input data, it generates multiple physically plausible, high-resolution precipitation maps,” said Luca Glawion of IMK-IFU, who developed the model while working on his doctoral thesis in the SCENIC research project. “Details at a resolution of 2 kilometers and 10 minutes become visible. The model also provides information about the statistical uncertainty of the results, which is especially relevant when modeling regionalized heavy rainfall events.”

He also noted that validation with weather radar data from the United States and Australia showed that the method can be applied to entirely different climatic conditions.

Correctly assessing flood risks worldwide

With their method’s global applicability, the researchers offer new possibilities for better assessment of regional climate risks. “It’s the especially vulnerable regions that often lack the resources for detailed weather observations,” said Dr. Julius Polz of IMK-IFU, who was also involved in the model’s development.

“Our approach will enable us to make much more reliable assessments of where heavy rainfall and floods are likely to occur, even in such regions with poor data coverage.” Not only can the new AI method contribute to disaster control in emergencies, it can also help with the implementation of more effective long-term preventive measures such as flood control.

Elon Musk’s artificial intelligence chatbot, Grok, received a major update.

Musk introduced Grok 4 during a livestream on X late Wednesday, calling it “the smartest AI in the world.” He praised the chatbot’s capabilities, saying it is smarter than “almost all graduate students in all disciplines, simultaneously.”

Introducing Grok 4, the world’s most powerful AI model.

Watch the livestream with @elonmusk and the @xAI team now. https://t.co/Mjt6w21qwd

— Engineering (@XEng) July 10, 2025

“Grok 4 is at the point where it essentially never gets math/physics exam questions wrong, unless they are skillfully adversarial,” Musk said. “It can identify errors or ambiguities in questions, then fix the error in the question or answer each variant of an ambiguous question.”

RELATED STORY | Musk’s AI company scrubs inappropriate posts after Grok chatbot makes antisemitic comments

Musk, who also owns Tesla, said in a separate social media post that Grok will be integrated into the electric vehicles as early as next week.

Grok 4’s release came just one day after the earlier model, Grok 3, shared several controversial posts, including some that praised Adolf Hitler.

In a statement, xAI, the company behind Grok, said it is actively working to remove hate speech from the platform and took swift action to update the model.

The controversial posts have since been deleted.

RELATED STORY | X CEO Linda Yaccarino leaves social media platform after 2 years