A new Stanford Law School study found that AI models scored 86 percent accuracy on law examinations, above the 73 percent average of law professors. The results have reignited debates over the redefinition of legal reasoning in the AI era.
The International Mathematical Union (IMU) has issued a comprehensive warning about the impact of AI companies on academic mathematics research. According to Ars Technica, the mathematics community is voicing structural concerns about industry funding and data ownership.
Microsoft has introduced its first flagship AI model with advanced reasoning capabilities, MAI-Thinking-1, at the Build 2026 event. According to The Verge, the medium-sized model matches leading models on key software engineering benchmarks.
Data security firm Cyera is nearing a 12 billion dollar valuation in a 300 million dollar funding round led by Evolution Equity Partners. The 80x ARR multiple stands at extreme highs for the sector.
Ars Technica reports that dozens of Red Hat packages distributed through the company's official NPM channel were found to contain backdoors. Users who downloaded affected packages are advised to investigate immediately. The discovery has reopened the debate over software supply-chain security.
The Verge's hands-on review of Google's new '24/7' AI agent Gemini Spark says the tool 'can be shockingly good' at completing multi-step tasks in the background, but flags concerns over the financial cost of subscription tiers and the privacy trade-offs of giving the agent broad device access.
TechCrunch reports that Anthropic, the AI company behind the Claude model family, has filed paperwork with the US Securities and Exchange Commission for an initial public offering. The filing marks the first concrete step toward a public listing in a generation-defining test of AI sector valuations on Wall Street.
TechCrunch reports that Mach Industries, led by 22-year-old founder and chief executive Ethan Thornton, has reached a $1.8 billion valuation in a $300 million funding round — a four-fold jump in a year. The defense-tech company is developing five autonomous aerial systems and recently completed an acquisition of a defense-software company.
An analysis surfaced via Hacker News from the Financial Times opens the debate over whether weaker junior hiring in recent years should be attributed not to AI but to remote-work patterns.
According to Ars Technica, a startup is sending camera-equipped cleaners to households offering free cleaning in exchange for robot training data. The approach adds a new dimension to the debate over collecting AI data inside homes.
According to The Verge, Nvidia has introduced its RTX Spark chip with the aim of joining Intel, AMD, Apple and Qualcomm in the thin Windows laptop and mini-PC market. The company claims the chip is the 'most efficient PC chip ever built'.
According to TechCrunch, environmental activist Erin Brockovich is launching a new campaign to spotlight the local water and energy impacts of fast-spreading US data centres.
Ars Technica reports that the US Federal Communications Commission (FCC) under the Trump administration has warned all US radio and TV broadcasters that they could face an enforcement process similar to the early license-renewal procedure imposed on ABC. ABC said the FCC's decision is 'unprecedented' and 'has no legitimate purpose'.
The Verge reports that Microsoft is considering a possible criminal case against a security researcher who identifies as 'Nightmare Eclipse', on the grounds that the researcher did not follow 'proper coordination' in disclosing zero-day exploits. The case has reignited debate over responsible-disclosure practices in the sector.
TechCrunch reports that GitHub Copilot's transition from a fixed monthly subscription model to a token-based billing model has elicited 'what a joke' responses among developers. The new model aims to pass on AI usage costs to developers.
Japanese technology investment group SoftBank has announced an investment of up to €75 billion in France to expand artificial-intelligence data center capacity. The company aims to develop and operate up to 5 gigawatts of additional data center capacity, TechCrunch reports.
According to Ars Technica, Europol, the FBI and 12 national law-enforcement bodies have dismantled a massive botnet, called BadBox 3.0, comprising more than 17 million Android-based devices. The operation prevented approximately $3.2 billion of fraudulent traffic.
According to The Verge, Nvidia, the Microsoft Windows on Arm team and Arm Holdings have jointly teased the new N1X laptop processor at Taipei's Computex 2026. The chip brings Nvidia's AI accelerator technology to the PC market.
According to TechCrunch, Groq, which develops AI neural network processors, is seeking a new $650 million round following Nvidia's recent $20 billion 'not-acqui-hire' deal. The company's valuation could be around $8 billion in the current high-demand market.
According to TechCrunch, Elon Musk's space technology company SpaceX has won a $6.45 billion multi-year launch contract under the US Space Force's National Security Space Launch Phase 3 (NSSL Phase 3) framework. The decision is of strategic significance ahead of the company's planned IPO valuation review.
An investigation published by BBC Future shows that modern cars collect a wide range of personal data from users, including gas pump app usage, seat position and even voice command recordings. The article, discussed on Hacker News, says that privacy regulation in the auto industry needs to be strengthened at the federal level.
The U.S. Federal Bureau of Investigation (FBI) filed a complaint alleging that a Google engineer used internal search data to win about $1.2 million on the Polymarket prediction platform. The indictment said the engineer was able to know the outcomes of bets in advance.
Microsoft announced a redesign of Microsoft 365 Copilot, with the new build loading twice as fast as before and a 'progressive disclosure' interface that surfaces tools and controls based on the user's prompt.
New 'AV Tracker' data published by the state of Texas show that Waymo dominates the market with about 350 robotaxis, while Tesla trails far behind with 28 vehicles. The figures are the first official numbers released under the Texas Autonomous Vehicle Act enacted in 2025.
BBC News reports that the European Union has fined the China-based e-commerce platform Temu 200 million euros for failing to take adequate measures against the sale of illegal products. The decision stands out as one of the largest fines imposed under the EU's Digital Services Act (DSA).
Ars Technica reports that Nvidia has announced plans to invest 150 billion dollars a year to make Taiwan an AI 'epicentre.' CEO Jensen Huang's announcement is also read as a reflection of the unintended backfire of the Trump administration's plan to make the US the global AI hub.
The Verge reports that Anthropic launched its new flagship model Claude Opus 4.8 on Thursday, with the company emphasising training focus on 'honesty' and the model's ability to acknowledge mistakes and recover.
TechCrunch reports that Paris has emerged as the most important AI city outside Silicon Valley, driven by a maturing startup ecosystem and leadership from companies like Mistral AI. European founders are increasingly choosing to scale at home rather than relocate to the US.
Researchers have revealed a new attack technique by which websites, through the browser, can observe visitors' SSD disk activity to profile the user's device — a privacy concern beyond current browser protections.
Valve has raised the prices of its Steam Deck handheld gaming console models by more than $200; the increase reflects supply-chain pressures worldwide and uncertainty over tariffs.