scale ai launches deepseek while openai expands its platform

Key Takeaways

• The Trump administration is cracking down on Chinese companies accused of stealing U.S. AI technology, with a House committee advancing a bill to sanction foreign actors who steal key features of U.S. AI models.
• DeepSeek released preview versions of a new flagship AI model, calling it the most powerful open-source platform available, challenging OpenAI and Anthropic.
• Anthropic's Claude AI now integrates with 15 third-party services including Spotify, Uber, Uber Eats, Instacart, Resy, TurboTax, and more.
• Infor launched secure AI tools in its Velocity Suite and Infor Agentic Orchestrator to address data security and privacy concerns that prevent over half of businesses from scaling AI.
• The Europe AI in social media market is projected to grow from $0.96 billion in 2025 to $13.29 billion by 2034, a compound annual growth rate of 33.85%.
• Schools are slow to adopt AI, with some banning it, but a student argues AI should be regulated not banned as it is the future of all jobs.
• Researchers created a typology for AI's role in psychosis, identifying four roles: catalyst, amplifier, co-author, and object, but note AI as a direct cause is very rare.
• Using Apple CEO Tim Cook's simplicity rule for ChatGPT prompts—removing unnecessary extras—produces sharper and more inventive responses.
• Alaska Representative Nick Begich introduced the DATA Act to allow data centers to operate on self-contained power systems, leveraging the state's energy reserves for AI investments.
• Only 13% of EU businesses actively use AI as of 2024, but demand for personalized customer experiences is driving more AI integration.

Trump administration targets Chinese firms over AI model theft

The Trump administration announced plans to crack down on Chinese companies that it says are stealing U.S. artificial intelligence technology. A memo from science adviser Michael Kratsios accused foreign entities, mainly in China, of running large-scale campaigns to extract capabilities from American AI systems. The administration will work with U.S. AI companies to identify these activities and punish offenders. This comes as China narrows the gap with the U.S. in AI performance, with a Stanford report saying the gap has effectively closed. A House committee also advanced a bill to sanction foreign actors who steal key features of U.S. AI models. Chinese officials denied the allegations and said they oppose unfair suppression of their companies.

Trump administration vows crackdown on Chinese AI exploitation

The Trump administration is promising to crack down on foreign tech companies that exploit U.S. artificial intelligence models, specifically targeting China. Michael Kratsios, the president's chief science and technology adviser, issued a memo accusing foreign entities of stealing American AI expertise. The administration plans to work with U.S. AI companies to identify these activities and find ways to punish offenders. This action comes as China continues to close the gap with the U.S. in the AI race.

Europe AI in social media market to reach $13 billion by 2034

The Europe artificial intelligence in social media market was valued at $0.96 billion in 2025 and is expected to grow to $13.29 billion by 2034, with a growth rate of 33.85% each year. AI in social media uses machine learning and algorithms to improve user experience, content creation, and marketing on platforms like Facebook and Instagram. The Digital Services Act is pushing companies to adopt AI for content moderation, while the General Data Protection Regulation limits data use and requires privacy-focused AI solutions. Nearly all EU residents are online, with 93% using the internet in 2024, and 70% use e-government services. Only 13% of EU businesses actively use AI as of 2024, but demand for personalized customer experiences is driving more AI integration.

AI is changing teaching but slowly in schools

Artificial intelligence is becoming common in daily life, but schools have been slow to adopt it. A high school social studies teacher shares that many Michigan school districts are taking a cautious approach, with some banning AI use unless teachers give permission. Students are using AI for homework and sometimes to cheat, but detection software is not reliable. The teacher predicts that in the future, AI will give each student a personalized tutor, and teachers will become managers of learning plans rather than just instructors. However, there is a fear that this could weaken the special bond between teachers and students that comes from discussing class content together.

Researchers create typology to understand AI psychosis

Despite media reports about AI psychosis, doctors are not seeing a wave of patients with this condition in clinics. Dr. John Torous and his co-authors created a typology to better understand AI's role in psychotic symptoms. They identified four roles AI can play: catalyst, where AI triggers new symptoms; amplifier, where it worsens existing symptoms; co-author, where it encourages risky actions; and object, where it becomes the focus of delusions. The researchers caution that many media reports lack medical context and that AI as a direct cause of psychosis is very rare. They hope this work helps clinicians better understand and address the issue.

DeepSeek releases new flagship AI model one year after breakthrough

DeepSeek has released preview versions of a new flagship artificial intelligence model, calling it the most powerful open-source platform available. This launch comes one year after the Chinese startup shook Silicon Valley with its earlier breakthrough. The new model challenges competitors like OpenAI and Anthropic in the AI race.

Infor launches secure AI tools to help businesses scale safely

Research shows that more than half of businesses struggle to scale AI, with data security and privacy being the main concerns. Industry cloud provider Infor has introduced new capabilities in its Velocity Suite and Infor Agentic Orchestrator to address this gap. These tools offer industry-specific AI solutions built on secure and compliant architecture. The study found that 34% of businesses in the US and Germany, 32% in France, and 45% in the UK cite data sovereignty and security as barriers to AI implementation. Infor's CEO Kevin Samuelson says their agents have deep industry knowledge that generic AI cannot match, and security protocols are built into the architecture from the start.

Using Tim Cook simplicity rule improves ChatGPT prompts

A prompt engineer tested Tim Cook's simplicity philosophy on ChatGPT prompts and found it boosted creativity. The rule involves removing unnecessary extras and focusing on what really matters. Instead of long, complex prompts with many instructions, simpler prompts like 'Give me 10 fresh campaign ideas for a wellness brand aimed at busy parents' produced sharper and more inventive responses. The engineer found that too many rules narrow possible outcomes, while simplicity leaves space for surprising connections. This approach mirrors Apple's design culture of clarity and focus.

Student says AI is the future and schools must embrace it

A high school student from San Diego wrote an opinion piece arguing that AI use in schools should be regulated but not banned. The student says teachers use AI as a scapegoat for cheating, but cheating has always existed and methods have just changed. The student believes AI is necessary for the incoming generation because it is the future of all jobs. Educators should not continue to associate AI use with cheating and should instead embrace it as previous generations embraced new technology.

Claude AI adds connectors for Uber Spotify TurboTax and more

Anthropic announced that its Claude AI now integrates with many third-party services including Spotify, Uber, Resy, and others. These connectors allow Claude to help with tasks like ordering food through Uber Eats and Instacart, making restaurant reservations via Resy, and starting tax preparation with TurboTax. Spotify integration lets Claude recommend music and podcasts and control playback. The full list of new connectors includes AllTrails, Audible, Booking.com, Instacart, Intuit Credit Karma, Intuit TurboTax, Resy, Spotify, StubHub, Taskrabbit, Thumbtack, TripAdvisor, Uber, Uber Eats, and Viator. All these integrations are available now, with more expected to come.

Alaska can use energy advantage to win AI data center investments

Representative Nick Begich argues that Alaska has a unique opportunity to capitalize on the AI boom because artificial intelligence requires massive amounts of energy. While other states face limits on power availability, Alaska has vast reserves of natural gas and untapped renewable energy. Begich introduced the DATA Act, which would allow large-scale energy users like data centers to operate on self-contained power systems that do not connect to the broader grid. This approach protects ratepayers from higher costs and ensures grid stability. Alaska's size and proximity to resource development make it well-suited for this model, and the state must act now to attract investment before other states win the competition.

Sources

• Trump administration vows crackdown on Chinese firms 'exploiting' U.S. AI models
• Trump administration vows crackdown on Chinese companies 'exploiting' U.S. AI
• Europe Artificial Intelligence in Social Media Market Size, Share & Trends, 2034
• In the age of AI, teaching may never be the same
• What to make of ‘AI psychosis’?
• DeepSeek Unveils Flagship AI Model a Year After Breakthrough
• Infor: Crossing the Data Security Barrier in Scaling AI
• I applied Tim Cook’s ‘Simplicity Rule’ to my ChatGPT prompts — and it became a masterclass in creativity
• A Student’s Perspective on AI in Schools (Opinion)
• Claude’s new connectors use AI to order food, control music, and do your taxes
• Opinion: How Alaska can use our energy advantage to capitalize on the AI gold rush