CrowdStrike reports Chinese hackers used AI-powered ARTEX framework to breach over 200 organizations globally, including foreign ministries, defense firms, and telecom providers, since October 2026.
14h ago
TD SYNNEX, Ingram Micro, and Westcon-Comstor reported double-digit sales growth driven by AI infrastructure, cybersecurity, and cloud spending, though margin pressures are reshaping distribution economics toward services.
14h ago
Shield Technology's AI benchmarks show top models resolve only 61.5% of complex IT tickets, highlighting gaps in following customer instructions and forcing MSPs to reconsider automation strategies and technician roles.
15h ago
Meta's Muse shopping assistant shows minimal corporate impact despite adoption claims. Companies across 50 firms report negligible orders, though many are optimizing websites preemptively for future agentic commerce growth.
17h ago
Amazon's September memo on AI usage has sparked engineer anxiety over whether documenting AI ineffectiveness will be career-limiting, with contradictory guidance allowing managers flexibility in performance decisions.
18h ago
Amazon CEO Andy Jassy predicts six to eight competitive AI models will emerge within three years, each strong enough for most workloads, driving multi-model adoption and positioning Bedrock as the neutral platform.
18h ago
University of Manchester researchers find AI adoption hasn't yet delivered forecast productivity gains despite widespread use, citing upfront costs and gaps between task-level improvements and organization-wide outcomes.
1d ago
Amazon CEO Andy Jassy predicts AWS' Bedrock AI service could match EC2's scale as businesses adopt multiple models, with token volume up 17-fold year over year.
1d ago
OpenAI's GPT-6.1 Terra runs three times faster and costs a tenth of GPT-6 Astra but scored lower on numerical reasoning, analysis finds.
1d ago
A new free framework lets security teams manually assess AI system ransomware vulnerabilities on a zero-to-five maturity scale, mapping controls to NIST and CISA standards without uploading internal data.
1d ago
Gartner's 2026 survey of 697 AI leaders shows 67% find bias monitoring difficult as CIOs struggle to explain algorithmic decisions to auditors, revealing traditional frameworks can't audit unpredictable AI systems.
1d ago
China's Volt Typhoon used Silent Train, a covert network of compromised devices, to exploit zero-day flaws and maintain strategic access to hundreds of U.S. critical infrastructure targets.
1d ago
Muse Technologies, founded by former NASA CIO Dr. Linda Cureton, launches defense contractor service addressing documentation failures that have cost companies millions in cybersecurity compliance settlements.
1d ago
Anthropic's engineering manager ended coding interviews, saying AI outperforms humans, but ZipRecruiter data shows 62.5 percent of tech employers still require formal skill tests on applications.
1d ago
Google and NASA are developing a robotic medical assistant for deep-space missions where communication delays up to 44 minutes make real-time Earth guidance impossible for astronauts.
1d ago
Anduril now commits its 2027 laptop will meet the 65% US component threshold, reversing its initial uncertainty after two days of questions about domestic manufacturing claims.
2d ago
Bluesky blocks links to leaked archive of over 41,000 DHS employee records, labeling them malicious despite transparency guidelines. Company cites unauthorized personal information sharing as violation.
2d ago
Cloudflare launched a beta Web Search API via AI Gateway to stop AI agents from guessing URLs and hitting 404 errors, delivering live results from three providers with unified billing.
2d ago
Datadog's Scout model investigates production alerts at 25% the cost of Claude Fable 5.1. Built on GLM-5.3 Flash, it runs inside Bits Investigation, diagnosing failures before engineers respond.
2d ago
A factory's senior IT manager used a profanity-filled password meant to insult hackers, but a security tester gained full network control through an unpatched Windows 7 server still connected to the domain.
2d ago
Cadence's Maestro AI portfolio claims 30% better chip design results and five times token efficiency versus general models. Early customer access begins fourth quarter 2026. Approach signals enterprise shift toward domain-specialized AI deployment.
2d ago
NIST released guidance for federal agencies to adopt post-quantum cryptography by 2031 deadlines mandated by Trump's executive order, urging immediate action and risk-based prioritization amid warnings quantum threats could emerge sooner.
2d ago