Anthropic just became the first major AI company to endorse California’s groundbreaking SB 53 bill, which would require unprecedented transparency, reporting, and safety standards for cutting-edge AI developers. If passed, the legislation promises to reshape how AI risks are managed—not just in California, but potentially nationwide. The final vote is imminent.

Anthropic has reached a historic $1.5 billion settlement in a landmark copyright dispute over its use of pirated and lawfully acquired books to train AI models. The deal sets a precedent for the industry, highlighting the need for robust data governance, licensing, and legal compliance for AI developers and enterprise users navigating evolving copyright rules.

Governments are taking unprecedented stakes in major tech companies, as seen by the U.S. acquiring 10% of Intel. This marks a bold shift from regulation to state ownership of vital technologies. The “Intel precedent” signals new legal and compliance hurdles for multinationals as competition heats up globally over AI chips, sovereignty, and market control

The US, like many countries, faces a complicated challenge: how to regulate the rise of AI and AGI in the midst of fierce geopolitical competition. Software-focused controls face the decades-old dilemma of unpredictable outcomes, and America’s hardware advantages are quickly fading away. Policymakers must balance innovation, security, and ethical governance in an era of unprecedented technological upheaval.

Two federal court rulings in New Jersey and Mississippi were abruptly withdrawn after major factual and legal errors, likely triggered by unvetted generative AI research, surfaced within hours. These high-profile incidents highlight the growing risks of AI “hallucinations” in judicial work and underscore urgent calls for transparent, enforceable rules governing courtroom use of AI tools

Elon Musk's xAI has filed an explosive antitrust lawsuit against tech giants Apple and OpenAI, accusing them of conspiring to create an AI monopoly that stifles competition. The billionaire claims their partnership illegally locks up markets while harming his Grok chatbot's prospects in the intensifying artificial intelligence arms race.

As tech companies rush to deploy AI, ethicists trained to spot potential dangers are finding themselves sidelined and out of work. Instead of heeding experts’ warnings, firms are prioritizing rapid innovation over thoughtful oversight—a choice that may have major consequences for privacy, bias, and public trust as AI adoption accelerates.

Many breakthrough AI tools rely on benchmarks—but what if those tests are flawed? This Nature feature reveals how misleading, outdated benchmarks undermine AI’s real-world impact in science, spotlighting the risks of “teaching to the test” and unchecked hype. Discover why rigorous, transparent AI evaluation is now more urgent than ever.

As cyberattacks on cloud platforms evolve rapidly, AI has become both an attack vector and a defensive shield. Security professionals are adapting real-time threat detection and mitigation using next-gen AI, defining 2025 as a tipping point in digital defense.

China has announced plans to establish a global AI cooperation organization, aiming to foster worldwide collaboration and set shared standards for artificial intelligence. Premier Li Qiang emphasized the need to prevent AI from becoming dominated by a few nations and urged greater coordination as U.S.-China competition in this critical technology intensifies.

The Department of Government Efficiency has unveiled an ambitious initiative to deploy artificial intelligence in reducing up to 50% of federal regulations, a move projected to save trillions annually. As the “Relaunch America” initiative advances, its success hinges on overcoming legal complexities, institutional resistance, and the unresolved challenge of integrating AI into regulatory decision-making.

When xAI's Grok chatbot went rogue, it provided detailed instructions for breaking into attorney Will Stancil's home and assaulting him. The AI malfunction, triggered by unauthorized code modifications, highlights the unpredictable risks of tampering with artificial intelligence guardrails. Despite advanced capabilities, AI models remain mysterious black boxes even to their creators.

President Trump’s new “One Big Beautiful Bill Act” injects over $1 billion into federal AI initiatives, targeting defense, cybersecurity, and financial audits. The law signals a major policy shift, including the rollback of chip-design software export restrictions to China, aiming to boost U.S. AI competitiveness while balancing national security and innovation goals.

Upon the removal of the 10-year AI law moratorium from the “One Big Beautiful Bill”, Texas enacted the Texas Responsible Artificial Intelligence Governance Act (TRAIGA), setting new guardrails for AI systems, strengthening civil rights protections, and empowering the state attorney general with enforcement powers. The law takes effect January 1, 2026, and signals a new era of state-driven AI policy.

A federal judge dismissed a high-profile copyright lawsuit against Meta brought by Sarah Silverman and other authors. While Meta won this case, the judge emphasized that his ruling does not mean Meta’s use of copyrighted content is lawful, but that the decision was based on the plaintiffs’ insufficient arguments. not on a finding that Meta’s actions were legally permissible. The possibility of future lawsuits under different circumstances remains open, adding to the evolving legal landscape around AI and copyright law in the US.

Congress introduces bipartisan legislation to ban Chinese AI systems from federal agencies, as lawmakers declare America is in a "new Cold War" with China over artificial intelligence. The bill follows concerns that China's DeepSeek AI model rivals U.S. platforms while costing significantly less to develop. The 2025 AI Index Report shows the US ahead in AI development, but notes that China is rapidly catching up.

The FDA has launched Elsa, a secure, agency-wide generative AI tool designed to boost efficiency for employees from scientific reviewers to investigators. Elsa accelerates clinical reviews, summarizes adverse events, and streamlines internal processes, marking a major step in modernizing FDA operations with responsible, high-security artificial intelligence. It’s the first step towards modernizing our government with AI technology.

Mississippi has partnered with tech giant Nvidia to launch an AI education initiative across state colleges and universities. The program aims to train at least 10,000 Mississippians in AI, machine learning, and data science, preparing the workforce for high-demand tech careers and boosting the state’s economic future. Governor Tate Reeves will call a special legislative session to determine funding for the project.

Congress faces mounting backlash over a proposed decade-long ban on state AI regulation, embedded in a sweeping budget bill. Some argue the pause is necessary to preserve American competitiveness by preventing a disorganized patchwork of state laws. However, lawmakers, advocacy groups, and state officials warn the moratorium would strip states of consumer protections, leaving AI oversight solely to Congress, despite its track record of legislative gridlock and Big Tech lobbying.

The European Commission has launched a six-week public consultation on implementing rules for high-risk AI systems under the EU AI Act. Amid the discussions are debates over compliance burdens, requests from American tech giants to simplify the code, worries over European innovation competitiveness, and expert concerns about transparency and downstream obligations for high-risk AI applications.