AI
Malaysia targets $5B in GDP gains from AI annually: Communications Minister
Malaysia is targeting as much as MYR20 billion ($4.7 billion) a year in additional economic output from artificial intelligence (AI) by 2030 and wants to shift from building data centres to applying AI across its economy.
Meta launches Muse personal AI agent in the US
Meta has launched Muse, a personal AI agent that can work across connected apps, continue tasks in the background and seek approval for sensitive actions. It is rolling out first in the United States through a dedicated app and WhatsApp.
Oracle plans global HPE Juniper rollout for AI data centers
Oracle plans to expand HPE Juniper routing, switching and telemetry across its global AI data center network, though commercial terms remain undisclosed.
Australia proposes social media algorithm opt-out
Australia has proposed requiring social media platforms to offer users a choice between personalized recommendations and a feed limited to accounts they follow, alongside a broader digital duty of care for online services.
AI attacks are faster, but defensive choke points remain, says Ensign InfoSecurity’s Teo Xiang Zheng [Q&A]
Ensign InfoSecurity’s Teo Xiang Zheng explains where frontier AI models still struggle during cyberattacks and which controls continue to create defensive friction.
AI is only as valuable as the discipline behind it
Adobe JAPAC President Ben Goodman argues that AI value depends on how well organizations connect data, content and workflows while maintaining trust and operational discipline.
Crealights, GlobalFoundries deepen silicon photonics partnership for 6.4T AI interconnects
Crealights and GlobalFoundries are expanding their silicon photonics collaboration around 6.4T near-package optics for high-density AI infrastructure.
Check Point finds now-closed ChatGPT flaw that enabled cross-account data access
Check Point Research disclosed a now-closed ChatGPT security issue that created a covert cross-account channel and could expose connected-app data.
Meta, Panmnesia propose CXL architecture for AI data centers beyond the rack
Meta and Panmnesia researchers have outlined a CXL-based architecture intended to extend tightly coupled AI computing across more servers and racks.
Palantir names Nebius preferred sovereign AI infrastructure partner
Palantir has named Nebius its preferred sovereign AI infrastructure partner and plans to integrate Nebius compute and inference services into its enterprise environment.
