Vals AI tested language models on China-related questions and found results varied by model, prompt language and format.
Chinese AI Models Are Politically Complex, Not Just Government Mouthpieces, Study Finds
Browsing Category
Politics & Government
177 posts
The Price Of Limited Detection: Rhysida’s Attack On Berlin’s Administration
Berlin says Rhysida stole about 1.4 million files. An rbb24 investigation found the detection system covered just 5% of devices in one department.
Fair Use Benefits Everyone
OpenAI has a page titled “Fair use benefits everyone,” but its publication date, argument and supporting evidence could not be verified.
Two Days After The Patch: Tracing PHP2Shell To Root On Palo Alto Servers
Dream reports an attacker rooted 24 Palo Alto firewalls and reached a government-affiliated network after PHP2Shell fixes were announced.
DeepSeek’s Efficiency Reputation Meets A Reported $16 Billion Raise
Reports say DeepSeek closed a 100 billion-yuan funding round, but the company has not confirmed the amount, investors or terms.
A Guide To Cisco SD-WAN Malware And Attacks On Network Edge Devices
Cisco Talos says KeyDrift malware stole SD-WAN key material after exploiting a critical flaw. Separate reporting links the flaw to Volt Typhoon.
Britain’s AI Testing Costs Outpace The US, But By Less Than 10 Times
Britain is adding £120 million to its AI Security Institute over four years. A fair annual comparison puts UK funding about eight to nine times US funding.
A Look At Europe’s Government AI Use: Jobs, Asylum And Bot Replies
A report examines Dutch public-sector job cuts, UK asylum AI summaries and bot submissions, and Sweden’s restrictions on public AI tools.
Expanding Our Partnership With GSA To Bring Claude To All Levels Of Government – Anthropic
Anthropic says it is expanding its GSA partnership to bring Claude to government, but has not disclosed coverage, terms, safeguards or a rollout date.
AI Policy Showdown: Federal Calls For Less Regulation, Virginia’s Cyber Order
Virginia’s Oct. 23 directive calls for a statewide plan against cyberattacks as the federal government presses for a lighter national AI framework.
