House panel to examine who pays for data centers’ water demands
A House hearing will examine draft measures on drinking-water systems and data-center costs, including whether developers or households should pay for upgrades.
A House hearing will examine draft measures on drinking-water systems and data-center costs, including whether developers or households should pay for upgrades.
Marvell reported record revenue and raised its forecasts, but shares fell as investors focused on when its expanded Google custom-chip deal may add revenue.
A federal grid-security order, FERC proceedings and Texas oversight are colliding with local concerns about AI data centers, power costs, water and land.
An Authors Guild survey found reported registration failures on 804 books, but the self-selected results do not prove industrywide breaches or individual wrongdoing.
A federal judge ruled the Pentagon unlawfully retaliated against Anthropic, but the decision does not require the military to resume using Claude.
A new OPM memo urges agencies to use AI in federal hiring while requiring documented controls when automated outputs could affect who advances.
SK hynix broke ground in West Lafayette on a more than $4 billion AI-memory packaging project, but production is not expected until 2029.
A new National Archives memo tells agencies when AI prompts, outputs and audit trails may need preservation, while stressing that records are not automatically public.
A Massachusetts judge allowed Universal and Sony to add a DMCA claim alleging Suno bypassed YouTube protections to obtain recordings for AI training.
The FTC finalized three settlements with Cox Media Group and two firms over allegedly deceptive claims about voice data, consent and local ad targeting.
New BLS projections show office-support jobs declining over the next decade while software, data and cybersecurity roles grow, but not as an immediate layoff forecast.
A federal judge found the Pentagon unlawfully punished Anthropic over AI safety limits, while leaving agencies free to choose other vendors through lawful procurement.
Australia’s ARIA will exclude wholly AI-generated recordings from its charts and awards, while human-led music using AI in limited roles remains eligible.
Google’s Expert Intelligence lets buyers use eligible e-books in Gemini Notebook, adding a licensed-use model as publishers challenge Google’s alleged AI training.
Nvidia posted $96.2 billion in fiscal second-quarter revenue, beat Wall Street expectations and forecast about $108 billion next quarter.
The BLS projects 5.9 million more U.S. jobs by 2035, with growth in technical fields and declines in office work—but AI exposure is not a layoff forecast.
NERC is taking comments through September 18 on draft reliability standards for AI data centers, crypto facilities and other large computational loads.
Emerald AI’s $150 million Series A signals investor interest in software that could help data centers manage power demand as U.S. grid constraints grow.
Indonesia is refining a copyright bill on AI-assisted works, training data, platform duties and digital royalties, but the proposal is not yet law.
A Senate hearing examined personalized pricing, electronic shelf labels and consumer-data use while stressing that no nationwide grocery-pricing rule has changed.
Hachette, Cengage, Elsevier and Scott Turow accuse Google of copying books and journals for Gemini training, raising questions about licensing and author pay.
APEC’s July 23 Chengdu Statement links AI security, connectivity, privacy, scam prevention and digital skills, but creates no binding regional rules.