June 5, 2025

news  
Read News
AI News: Meta, EU AI Act, China AI, FTC, Dataiku

Several developments in the field of Artificial Intelligence (AI) are unfolding across various sectors. Meta has entered a 20-year agreement to purchase 1.1 gigawatts of nuclear power from Constellation's Clinton Clean Energy Center starting in 2027 to power its AI and data centers, highlighting the significant energy demands of AI and prompting other tech companies like Amazon, Google, and Microsoft to explore new energy sources. Meanwhile, the European Commission has released guidelines clarifying the EU Artificial Intelligence Act (AI Act), which took effect in February 2025, outlining prohibited AI practices. China is leveraging AI and robotics to enhance its supply chains, with factories increasingly adopting these technologies to cut costs and improve quality. In the U.S., the Federal Trade Commission (FTC) is challenging companies making false AI claims, while the California Senate has passed a bill to ensure AI chatbot safety, particularly for young users. Dataiku has been recognized as Snowflake's 2025 AI Data Cloud Product Partner of the Year for its AI integration and support. Pennsylvania lawmakers are advocating for states' rights to regulate AI, and Garfield AI, an AI-based legal firm, has gained approval to manage small claims with minimal human intervention. Experts like Yoshua Bengio are focusing on AI safety, with Bengio launching LawZero, an AI nonprofit, to monitor AI systems. Concerns are also being raised about the potential impact of AI on human identity and the importance of prioritizing safety over profit in AI development.

Key Takeaways

  • Meta is investing in nuclear power to meet the energy demands of its AI and data centers.
  • The EU has released guidelines clarifying prohibited AI practices under the AI Act.
  • Chinese factories are increasingly using AI and robotics to improve supply chains.
  • The FTC is challenging companies making deceptive AI claims.
  • California is working on legislation to make AI chatbots safer, especially for young people.
  • Dataiku was named Snowflake’s 2025 AI Data Cloud Product Partner of the Year.
  • Pennsylvania lawmakers are asserting states' rights to create AI laws.
  • An AI legal firm, Garfield AI, has been approved to manage lawsuits with limited human assistance.
  • Experts are raising concerns about AI's potential impact on human identity.
  • Yoshua Bengio launched LawZero, an AI nonprofit focused on AI safety.

Meta buys nuclear power for AI in 20-year deal

Meta will buy 1.1 gigawatts of power from Constellation's Clinton Clean Energy Center starting in 2027. The 20-year deal will power Meta's AI and data centers. Meta says nuclear energy is reliable for AI advancement. The partnership preserves over 1,100 jobs and provides $13.5 million in annual tax revenue. The Clinton plant was previously set to close but was saved by the Future Energy Jobs Act.

Meta's nuclear power deal shows AI's big energy needs

Meta's 20-year deal with Constellation Energy shows how much energy AI needs. Other companies like Amazon, Google, and Microsoft are also looking for new energy sources. AI uses a lot of energy, often from fossil fuels, which causes climate change. Meta is also planning to use natural gas for its data center in Louisiana. Renewable energy sources make up about 24% of data center power in the U.S., while nuclear is about 15%.

EU releases guidelines on prohibited AI practices under AI Act

The European Commission released guidelines in February 2025 to clarify the EU Artificial Intelligence Act (AI Act). The guidelines explain definitions and prohibited AI practices that took effect on February 2, 2025. The guidelines cover eight prohibited AI practices in Article 5 of the AI Act. They address the relationship between high-risk AI systems and prohibited practices. The guidelines clarify rules on personalized ads, lawful persuasion, and exploitation of vulnerabilities.

EU releases guidelines on prohibited AI practices under AI Act

The European Commission released guidelines in February 2025 to clarify the EU Artificial Intelligence Act (AI Act). The guidelines explain definitions and prohibited AI practices that took effect on February 2, 2025. The guidelines cover eight prohibited AI practices in Article 5 of the AI Act. They address the relationship between high-risk AI systems and prohibited practices. The guidelines clarify rules on personalized ads, lawful persuasion, and exploitation of vulnerabilities.

AI boosts China's supply chains giving them a bigger advantage

Chinese factories are using AI and robotics to cut costs and improve quality. Startup Cybord uses AI to find flaws in manufacturing components. China installed seven times more robots than the U.S. in 2023. Beijing plans to further digitalize factories by 2030 using AI and blockchain. Chinese companies are leaders in using digital technology in manufacturing. BYD is using robots for logistics and investing in robotics companies.

FTC challenges companies making false AI claims

The Federal Trade Commission (FTC) is taking action against companies making deceptive claims about artificial intelligence.

California Senate passes bill to make AI chatbots safer

The California Senate passed a bill to make AI chatbots safer, especially for young people. The bill requires chatbot platforms to remind users every three hours that the characters are not human. Platforms must also disclose that chatbots may not be suitable for minors. They must also have a plan for dealing with users who express thoughts of suicide or self-harm. The bill now goes to the California State Assembly.

Dataiku wins Snowflake AI Data Cloud Product Partner of the Year

Dataiku was named Snowflake's 2025 AI Data Cloud Product Partner of the Year. The award recognizes Dataiku's integration with Snowflake and its support for AI. Dataiku launched AI Agents with Dataiku, which supports the full lifecycle of enterprise AI agents. Dataiku connects to Snowflake Cortex AI services at every level. Customers can now buy Dataiku licenses directly through Snowflake Marketplace.

Pennsylvania lawmakers assert states' rights to make AI laws

Pennsylvania lawmakers are working on bills about artificial intelligence and pushing for states to have the power to create their own AI laws.

AI legal firm gets approval to manage lawsuits

Garfield AI, an AI-based legal firm, received approval to handle small claims with little human help. The AI can create letters, draft claims, apply for judgments, and handle settlements. Users upload information and authorize each step. The company says it reduces errors and delays. The Solicitors Regulation Authority (SRA) calls it a landmark moment and encourages more AI legal services.

Will AI make humans feel less important?

Tyler Cowen and Avital Balwit wrote an article about how AI will change humanity. They say AI will cause an identity crisis because humans will no longer be the smartest. People may struggle to find meaning in a world where AI is more capable. The authors still lean towards optimism about the future.

AI pioneer Yoshua Bengio wants to make AI safer

Yoshua Bengio, a leading AI expert, launched LawZero, an AI nonprofit focused on AI safety. LawZero will create Scientist AI, a system to monitor other AI systems. Scientist AI will explain the world from observations instead of imitating humans. Bengio is concerned about AI's impact on society if it's not controlled. He says current AI models are developing dangerous abilities. He believes AI should be designed for safety over profit.

Sources

AI Artificial Intelligence Energy Consumption Data Centers Nuclear Power Renewable Energy Meta Constellation Energy EU AI Act AI Regulation Prohibited AI Practices China Manufacturing Robotics Supply Chains FTC False AI Claims California AI Chatbots AI Safety Dataiku Snowflake AI Data Cloud Pennsylvania State AI Laws AI Legal Services Garfield AI AI and Identity Yoshua Bengio LawZero AI Safety Research