Sophia Andreou

Instagram will soon charge for fuller AI access, as confirmed by Adam Mosseri. Free daily caps remain, but a paid tier is in the works. The move follows Meta's massive AI infrastructure spending and a recent privacy backlash.

A Peking University team has created an all-optical link for standard chips, achieving a 100x speedup in distributed AI inference with drastically lower resource use. The breakthrough could reshape data center design amid soaring AI demand.

The UN's first scientific assessment of AI reveals a widening compute divide and dangerous corporate concentration, urging immediate policy action. Co-chaired by Yoshua Bengio and Maria Ressa, the report calls for global cooperation to prevent irreversible harm.

Despite rapid progress, frontier AI models still hallucinate with alarming confidence. This article dissects recent embarrassing failures, explains the underlying mechanisms, and offers practical mitigation strategies.

The Punjab Curriculum Board has approved AI as a subject for primary students, with labs in all schools. The rollout begins in 2027-28, but challenges like teacher training and rural connectivity loom large.

Anthropic appoints former Federal Reserve Chair Ben Bernanke to its independent oversight trust, signaling a serious focus on AI's economic disruption. His crisis experience aims to guide responsible AI development for long-term societal benefit.

Harvard psychiatrist Arthur Kleinman argues that as AI becomes embedded in healthcare, the moral practice of care risks being sidelined by efficiency metrics. He calls for systems that integrate technology without losing the human presence essential to healing.

Quantitative hedge funds suffered their worst performance in nearly a year after a sharp unwind in crowded AI and semiconductor trades, erasing a quarter of their 2026 gains. Goldman Sachs reported that systematic managers fell to 10.8% year-to-date returns from 14.4% in just weeks.

Enterprises that embed third-party AI into their workflows are learning that outsourcing the technology doesn't outsource the legal risk. From FTC actions to EU regulations, the deployer is now the primary target when models fail.

China is preparing to allow Alibaba, ByteDance, and DeepSeek to import Nvidia H200 chips, ending a months-long stand-off. The move acknowledges domestic chips can't yet meet AI computing demands.

A coalition of newspapers led by The New York Times has asked a federal court to sanction OpenAI, alleging the company lied about its ability to search for copyrighted content in ChatGPT and deleted billions of relevant conversations. The motion marks a dramatic escalation in the high-stakes copyright battle over AI training data.

Hong Kong's IPO market is splitting in two: small-cap debuts are crashing while the city bets big on AI giant Zhipu AI. This week's bloodbath saw Luxshare and others sink, with a major listing canceled despite record demand.

The French competition authority's probe into Nvidia over alleged anti-competitive practices is nearing completion, a top official said. The outcome could lead to massive fines and reshape the AI chip market.

Chinese technology companies, including AI model developers and chipmakers, have raised a combined $17.38 billion in Hong Kong equity markets this year, highlighting the city's role as a capital hub despite geopolitical headwinds.

Filmmaker Aranya Sahay argues that AI functions as a modern form of colonization, shaped by Western minds. His debut film, Humans in the Loop, explores this bias through an Adivasi woman's story, now streaming globally on Netflix.

Asian fund managers are moving beyond AI hype, targeting companies that can weather disruption and provide essential infrastructure. At Reuters NEXT Asia, leaders from Temasek, Goldman Sachs, and others detailed a cautious strategy focused on hard assets and real cash flows.

China is considering limits on foreign access to its most advanced AI models, a decision that could isolate major startups like Zhipu and Moonshot. This move mirrors U.S. restrictions and may accelerate global AI fragmentation, forcing other nations to build costly alternatives.

British Columbia is suing OpenAI, alleging the company failed to alert police to violent ChatGPT threats before a mass school shooting in Tumbler Ridge. The case could set a precedent for AI platform liability.

South Korean memory chipmakers Samsung and SK Hynix saw sharp declines on July 8, mirroring a U.S. semiconductor selloff. Investor concerns over the sustainability of AI-driven demand are shaking the market's foundations.

India’s engineering education faces a crisis as AICTE closes 58 colleges and cuts 950 courses. The AI revolution demands a shift from quantity to quality, but outdated curricula and weak regulation threaten the country’s tech ambitions.

Meta revealed that its AI systems removed 4 million accounts and 36 million pieces of content linked to child exploitation in early 2026. The disclosure follows India's demand for answers after CSAM ads appeared on Instagram.

Ukraine is moving to self-hosted AI models to ensure control over critical digital tools, reducing reliance on providers like OpenAI and Anthropic. The policy, driven by wartime needs and recent U.S. export restrictions, prioritizes on-premise deployments for government, business, and military applications.

Ant Group's AI unit, Robbyant, has developed LingBot-Vision, a model that detects object edges with sub-pixel precision. It outperforms larger models on depth benchmarks while using far less data, potentially reducing costly glass crashes in robotics.

Anthropic researchers have discovered a hidden internal workspace in Claude AI, called J-space, where the model silently processes concepts. The J-lens technique reads these thoughts, revealing safety insights and flexible reasoning.

The CEO of Abu Dhabi-based AI startup AppliedAI secretly pleaded guilty to insider trading, using tips from lawyers at major firms. The case, unsealed Monday, reveals a scheme that netted nearly $1 million in illegal profits.

Chinese authorities are discussing potential restrictions on overseas access to the country's most advanced AI models, including those still in development, according to sources. The move could reshape global AI collaboration and competition.

Hong Kong should craft a focused AI strategy leveraging its strengths in finance, healthcare, and urban development, experts said at a recent conference. The city's first five-year plan must avoid being too broad and instead prioritize competitive scenarios.

Josh Tyrangiel argues that AI adoption fails when companies chase technology instead of defining clear problems. He shares a Cleveland Clinic example where AI cut sepsis deaths by 41%, proving that domain expertise and disciplined execution matter more than hype.

Samsung Electronics forecasts a 19-fold profit surge in Q2 2026, surpassing three years of combined earnings, fueled by AI memory chip demand. However, shares fell as market skepticism grows over the sustainability of the boom.

Reddit is using AI to block 23 million spam views per day, catching 25,000 fake posts and comments daily. The crackdown aims to protect authentic conversations and the data that trains major AI models.