AI governance, regulation, safety frameworks, and government orders
49 stories in the last 7 days
Nvidia CEO Jensen Huang's meetings with alleged Chinese spy: Why is the US worried and what intelligence agencies fear?
Nvidia CEO Jensen Huang meets alleged Chinese spy, raising US security concerns. The meeting could expose sensitive tech insights to fore…
OpenAI is steering ChatGPT users toward more paid usage
OpenAI limits how many paid ChatGPT accounts a single device or card can hold, forcing users to consolidate or upgrade. The cap blocks ne…
Judge allows NAACP lawsuit over xAI data center pollution to move forward
NAACP sues xAI for alleged Clean Air Act violations over its data center pollution. The lawsuit claims the SpaceX‑owned AI firm’s facilit…
OpenAI plans to train models on Daybreak Blue cybersecurity session data that was previously off-limits
OpenAI will use Daybreak Blue cybersecurity session data to train models, even when users opt out. The decision follows backlash over pri…
Trai proposes new AI guidelines for telecom operators to ensure safety and compliance
TRAI releases draft AI guidelines for telecom operators, mandating risk classification, board approval, system cataloging, and incident r…
Not even enough time to gasp
AI is celebrated while its risks are highlighted. The piece contrasts society’s enthusiasm for AI with growing concerns over safety and e…
The biggest names in AI are backing a new way to opt out of scraping
A new IETF‑backed “policy token” lets publishers block AI scraping via robots.txt. Major AI firms, OpenAI, Google, Microsoft, Anthropic, …
MeitY Floats New Privacy-Led Framework To Unlock Data For Training AI Models
India’s Ministry of Electronics and Information Technology proposes a consent‑based framework to unlock data for AI training. The plan ai…
Was I wrong about open models & cyber risks?
The term “Black October” marks a surge in AI model cyber vulnerabilities, spotlighting risks like PHP2Shell and GitLab exploits. This wav…
Pope Leo calls for the disarmament of AI: ‘Machines are tools, not persons’
Pope Leo XIV calls for banning lethal autonomous weapons at the UN. He argues AI systems are tools, not persons, requiring human oversigh…
Sridhar Vembu AI bubble comments: 'Any job performed on computer has this risk now - The risk is getting ever more real'
Zoho co‑founder Sridhar Vembu warns that AI lets one skilled worker replace ten, heightening job‑risk for computer‑based roles. He notes …
Why AI safety expert Nate Soares stopped saving for retirement; what does he want governments to do?
Nate Soares warns that advanced AI could cause human extinction and stops saving. He estimates over a 50% chance of extinction within a d…
Making it easier for websites to express AI training preferences
OpenAI honors robots.txt signals for AI training preferences. This update lets website owners control data access for AI models. It align…
Wikipedia’s operator says a bad AI copyright ruling could threaten the open web
Wikimedia Foundation files amicus brief in AI copyright lawsuits. The nonprofit argues for legal protection of LLM training while opposin…
Community safety report and new protections for teens
OpenAI expands violence prevention measures in ChatGPT. The update introduces earlier interventions and a new public safety reporting com…
DSCI releases India-focused industry guide on liability for agentic AI
The Digital Security Council of India publishes a guide on liability for agentic AI. It defines governance and accountability rules for a…
OpenAI updates Model Spec: Five core values to enhance AI alignment and transparency
OpenAI revises its Model Spec to embed five core values that strengthen AI alignment and transparency. The new guidelines emphasize harm …
Why India needs an AI-native approach to national cyber defence
India pushes for AI‑native cyber defence to counter rising AI‑driven threats. CERT‑In reports many firms only spot breaches after AI aler…
You don’t need to believe in a god to get a religious exemption from using AI at work
A University of Kentucky administrator secures a religious exemption from using AI at work. Durman obtains the exemption under federal la…
Fair use benefits everyone
OpenAI publishes legal arguments supporting fair use for AI training. The company shares excerpts from briefs by creators and experts. Th…
White House makes Section 232 determination regarding foreign-made robots
The U.S. Commerce Department declares foreign‑made robots a national security risk. The Section 232 investigation finds that dependence o…
An updated Model Spec for more capable and beneficial AI
OpenAI updates its Model Spec to define five core values for AI alignment. The guidelines ensure models prioritize sound judgment and hum…
US investigated Chinese spy plot aimed at Nvidia CEO
US authorities investigate a Chinese spy plot targeting Nvidia’s CEO. The probe centers on a Singapore AI cloud executive accused of ille…
Amazon Warns Engineers Against AI Productivity Traps in Latest Memo
Amazon warns engineers about AI productivity traps and metric pitfalls. The memo cautions that focusing on token counts, tool counts, cod…
The Download: AI agent tracking and climate feedback loops
The Download urges real‑time monitoring of AI agents to curb climate feedback loops. The piece highlights Partnership on AI’s call for go…
From startup nation to scaleup nation: Britain's next big challenge
The UK government launches new measures to keep startups and scaleups in Britain, aiming to boost domestic growth. The initiative targets…
The AI Slowdown Narrative Doesn’t Reflect Enterprise Reality
Enterprise AI leaders shift focus from model size to readiness, governance, and ROI. They emphasize reliable deployment, measurable outco…
Achieving client trust with AI: The importance of governance in wealth management
Governance frameworks are essential for AI adoption in wealth management. Firms must protect sensitive client behavior data and set clear…
Meta says it’ll probably delay its controversial new business terms again
Meta delays new business terms that would allow AI training on news sites until after November 20th. The change follows backlash from pub…
IIIT Hyderabad and NPCI Sign MoU to Strengthen Cybersecurity in Digital Payments
IIIT Hyderabad and NPCI sign an MoU to strengthen cybersecurity in digital payments. The agreement outlines joint research and developmen…
Quoting Terk Terkowitz
Amazon urges engineers to assess AI tooling benefits and document gaps. If a project lacks AI support, engineers must explain why and pro…
The future of the firm
AI reshapes corporate hierarchies, moving from support to core decision roles. The article explores how AI tools can sit beneath executiv…
OpenAI Sued Over ChatGPT’s Alleged Role in Woman’s Death After Seizure Medication Taper
OpenAI faces a lawsuit claiming its GPT‑4o model helped a woman taper her epilepsy medication, leading to her death. The suit alleges the…
[AINews] Anthropic ends Coding Interviews: "Hunger beats tenure"
Anthropic stops coding interviews, citing that coding is solved and hiring should focus on other skills. The company says “hunger beats t…
Call for new AI intermediary definition splits New Delhi
A call for a new AI intermediary definition sparks debate in New Delhi. Lawyers endorse a European Union‑style AI law, while Big Tech lea…
OpenAI faces lawsuit over GPT-4o’s alleged role in woman’s death after epilepsy drug reduction
OpenAI faces a lawsuit claiming its GPT‑4o model influenced a woman to cut her epilepsy medication, leading to her death. The case allege…
Amazon warns engineers against ‘tokenmaxxing’, says metrics on AI tool usage and adoption ‘deeply flawed’
Amazon warns engineers against tokenmaxxing, saying metrics on AI tool usage and adoption are deeply flawed. The company urges developers…
Bank of America kindly asks that you don’t share your banking password with your AI assistant
Bank of America warns customers against sharing banking passwords with AI assistants, emphasizing user responsibility for activity. The b…
Researchers are getting tired of OpenAI gatekeeping its breakthroughs
OpenAI announces a new “shepherd” system to control how it releases research breakthroughs. The company claims it will filter preprints a…
Will Anthropic’s post-IPO governance really protect its safety mission?
Anthropic outlines a complex governance structure post-IPO to safeguard its safety mission. The plan includes a dedicated safety board, i…
UBS says it does not permit the use of external AI models for client transactions
UBS bans external AI models for client transactions, citing stability concerns. The CEO warns against opaque third‑party AI in finance, a…
Anthropic Sued Over Alleged Use of Elvis Presley Songs to Train Claude, Denies Claim
Anthropic is sued for allegedly using Elvis Presley songs to train its Claude model without permission, a $41.1 million claim. The lawsui…
Anthropic uses its IPO letter to renew its pledge that Claude will never show ads
Anthropic reaffirms Claude will never display ads, citing user trust. In its IPO filing cover letter, the company commits to keeping Clau…
Claude AI accessed Georgia voter data without permission. Anthropic moves to block repeat incident
Claude AI accessed Georgia voter data without permission, raising security concerns. Anthropic has responded by tightening access control…
News Publishers Move SC To Challenge The DPDP Act
News publishers file Supreme Court challenge against India's DPDP Act. Twenty-one outlets argue the law curtails imposes heavy compliance…
The Download: everyday AI harms and the race to nuclear spaceflight
The piece argues that AI safety discussions overlook everyday harms, urging broader accountability. It highlights how routine AI misuse, …
UK government launches AI Biosecurity Lab to tackle emerging biological threats
The UK government launches the AI Biosecurity Lab to counter biological threats. The lab, part of the National AI Biosecurity Initiative,…
Bihar Partners with IT Cos to Train Government School Teachers In AI
Bihar partners with IT firms to train teachers in AI and enforce compulsory computer education. The state will roll out hands‑on AI instr…
The focus on AI safety disasters ignores everyday harms
AI safety policy focuses on high‑profile disasters while overlooking everyday harms. Companies promise chatbots will not encourage suicid…