Google expands AI tools as Sundar Pichai faces worker concerns over Pentagon deal

Key Takeaways

Greece adds AI rule to new constitution to protect people

Greece is updating its constitution to ensure artificial intelligence serves human society and protects individual freedoms. Prime Minister Kyriakos Mitsotakis announced this change to safeguard future generations from AI risks. The new rule states that AI must support freedom and social prosperity while reducing dangers. Other changes include expanding voting options and increasing mandatory school years from nine to eleven. Experts say this legal step is needed because private tech companies hold too much power over data.

Greece adds AI rule to new constitution to protect people

Greece plans major constitutional changes to require artificial intelligence to serve human society and protect individual rights. Prime Minister Kyriakos Mitsotakis presented these revisions to ensure technology benefits future generations. The proposed law states that AI must support freedom and social prosperity while managing risks. The country also plans to expand postal voting and increase mandatory schooling from nine to eleven years. Experts argue this legal requirement is necessary because private tech platforms hold too much power.

Greece adds AI rule to new constitution to protect people

Greece is preparing constitutional changes that require artificial intelligence to serve individual freedom and social prosperity. Prime Minister Kyriakos Mitsotakis said these rules are needed to protect future generations from AI risks. The new law will ensure AI reduces dangers while fully realizing its benefits. Other updates include expanding postal voting and increasing mandatory school years from nine to eleven. Experts believe this legal step is vital because private tech companies hold too much power over data.

DeepMind workers vote to unionize over Pentagon AI deal

Workers at Google’s DeepMind lab voted to unionize after the company signed a deal with the U.S. Pentagon for AI projects. More than 600 employees signed a letter urging CEO Sundar Pichai to stop using AI in classified military settings. Workers fear AI could be used for lethal weapons or mass surveillance in authoritarian ways. The union wants stronger protections for whistleblowers and the right to refuse unethical work. Google says it supports national security but promises human oversight for AI use.

Pentagon promises to follow laws but AI limits remain unclear

The U.S. military uses artificial intelligence in drones and predictive analytics but faces legal and ethical questions. The Pentagon has promised to follow the law when deploying AI technology. However, critics say current guidelines are too vague and lack proper oversight. It is unclear who is responsible when an AI system makes a mistake or causes harm. The military must improve transparency and accountability to address these complex issues.

Ohio candidate who asked AI to praise Hitler loses primary

Ohio gubernatorial candidate Casey Putsch lost the Republican primary after asking an AI bot to list Hitler’s good qualities. Putsch received 17.5% of the vote while Vivek Ramaswamy won with 82.5%. Putsch made headlines for advertising a beer hall rally and using antisemitic language. His campaign rhetoric included dogwhistles that upset many voters. The election result shows the electorate rejected his controversial approach.

Report shares insights from inside China’s AI research labs

A recent report shares personal experiences from visiting AI labs in China during a trip from Hangzhou to Shanghai. The author described the journey as a warm and welcoming human experience despite the foreign setting. Many people in the AI ecosystem were met with humility and openness. The report highlights the growing presence of wind turbines and modern skyscrapers in the region. It emphasizes the human connections made within China’s advanced technology sector.

Healthcare CIO warns AI should not just optimize broken systems

Dr. Ash Goel, CIO of UCSF, warns that healthcare AI should not simply optimize existing inefficient processes. Speaking at the HIMSS Global Health Conference, he argued that squeezing margins from a broken system is not true innovation. Instead, AI must be used to improve patient outcomes and healing. The current system prioritizes throughput over care, and automating this dysfunction will only make problems worse. Healthcare leaders must ensure AI serves patients rather than just administrative efficiency.

SandboxAQ CEO says India is ready to lead AI era

Jack Hidary, CEO of SandboxAQ, told the World Economic Forum that India is positioned to lead the “AI or Die” era. He noted that 80% of India’s economy is in the real world, giving it a unique advantage. AI could compress drug discovery timelines for cancer and Alzheimer’s as FDA guidance matures in 2026. Cybersecurity is a top priority since the WEF ranks it as the sixth biggest global risk. Hidary believes AI offers both benefits and new vulnerabilities for banking and public utilities.

Gmail adds AI Inbox to mobile apps for better task management

Google is adding an AI Inbox feature to Gmail for Android and iOS devices. The new interface appears in the navigation drawer below Gmail and Chat with four tabs total. It shows suggested to-dos and topics to catch users up on important matters. The feature is currently in beta for Google AI Plus, Pro, Ultra, and Business users. It helps reduce the need to switch between different Google Workspace apps for tasks.

European translators worry about partnering with US cloud firms

European AI translator DeepL announced a partnership with Amazon Web Services to improve its global services. This move has raised concerns among European users about data privacy and US government access. Critics fear it strengthens Silicon Valley’s monopoly on digital infrastructure. DeepL says AWS will not access customer data or use it to train AI models. Some customers have already canceled subscriptions due to these privacy concerns.

Multiple AI models create routing challenges for big companies

Enterprises are facing difficulties managing multiple AI models in their production systems. A recent report found that 78% of organizations run their own AI inference services. Companies now operate an average of seven different AI models for various tasks. This creates complex routing and security challenges for IT teams. Organizations must decide which model handles each request based on cost, speed, and compliance needs.

Sources

• Greece, birthplace of democracy, seeks to put humanity ahead of AI in updated constitution
• Greece, birthplace of democracy, seeks to put humanity ahead of AI in updated constitution
• Greece proposes constitutional safeguards on artificial intelligence
• DeepMind Workers Vote to Unionize Following Google’s Pentagon AI Deal
• The Pentagon keeps promising to follow the law when using AI, but what are the limits?
• Ohio gubernatorial candidate who asked AI bot to praise Hitler is pummeled in GOP primary
• Notes from inside China's AI labs
• What healthcare AI cannot be allowed to become, according to one CIO
• SandboxAQ CEO Jack Hidary: India Positioned to Lead the "AI or Die" Era
• Gmail adds AI Inbox to the bottom bar as Help me write becomes more personal
• Europe’s AI translation industry told it risks reputation by partnering with US firms
• Multi-model AI is creating a routing headache for enterprises