New MLCommons benchmarks to test AI infrastructure performance

Posted on by Admin
New MLCommons benchmarks to test AI infrastructure performance

The latest release also broadens its scope beyond chatbot benchmarks. A new graph neural network (GNN) test targets datacenter-class hardware and is designed for workloads like fraud detection, recommendation engines, and knowledge graphs. It uses the RGAT model based on a graph dataset containing over 547 million nodes and 5.8 billion edges. Judging performance Analysts suggest that these benchmarks will make it easier to judge the performance of various hardware chips and clusters based on…

Read More

Want AI to work for your business? Then privacy needs to come first

Posted on by Admin
Want AI to work for your business? Then privacy needs to come first

Yuliya Taba/Getty Images Cisco has released a “2025 Data Privacy Benchmark Study” that looks at the privacy challenges companies face with the rise of artificial intelligence. It offers practical insights for businesses that want to integrate AI while keeping privacy front and center. Also: The best free AI courses and certificates in 2025 The study gathered opinions from 2,600 privacy and security professionals across 12 countries. A key finding is that most companies (86%) support privacy…

Read More

Over Half of Attacks on Electricity and Water Firms Are Destructive

Posted on by Admin
Over Half of Attacks on Electricity and Water Firms Are Destructive

Over three-fifths of US and British water and electricity firms were targeted by cyber-attacks in the past year, with a majority suffering serious disruption, according to Semperis. The security vendor polled IT and security professionals at 350 water treatment plants and electricity operators in both countries, to compile its report, The State of Critical Infrastructure Resilience. Of the 62% that said they suffered a cyber-attack in the past year, 80% were hit multiple times. Worse,…

Read More

Over 500 Phishing Domains Emerge Following Bybit Heist

Posted on by Admin
Over 500 Phishing Domains Emerge Following Bybit Heist

A large number of phishing campaigns emerged in the aftermath of the Bybit heist, designed to siphon cryptocurrency from its customers, according to BforeAI. The security vendor detected 596 suspicious domains originating from at least 13 different countries in the three weeks following news of the biggest crypto theft in history. Dozens of these domains spoofed the cryptocurrency exchange itself, many using typosquatting techniques and including keywords such as “refund,” “wallet,” “information,” “check” and “recovery.”…

Read More

한풀 꺾인 개발자 수요··· 좋은 시절은 지났을까?

Posted on by Admin
한풀 꺾인 개발자 수요··· 좋은 시절은 지났을까?

소프트웨어 개발자들이 전 세계를 지배하던 때가 있었다. 기본적인 코딩 경력이 있는 사람이라면 누구나 여러 가지 일자리를 얻을 수 있었던 시절이다. 그러나 이는 벌써 5~6년 전 이야기다. 개발자들은 더 이상 IT 고용 시장에서 ‘슈퍼 갑’이 아니다. 데이터 엔지니어, 머신러닝 엔지니어, AI 제품 관리자 같은 직책이 전통적인 소프트웨어 개발자의 수요를 대체했다. 기업들이 AI를 도입하기 위해 서두르고 사이버 보안 전문가를 물색한다. 새로운 현실의 예는 세일즈포스에서 찾을 수 있다. 2월에 마크 베니오프 CEO는 CNBC의 스쿼크 박스(Squawk Box)에서 2025년이 회사…

Read More

Google DeepMind Proposes AI ‘Monitors’ to Police Hyperintelligent Models

Posted on by Admin
Google DeepMind Proposes AI ‘Monitors’ to Police Hyperintelligent Models

Google DeepMind has introduced a new approach to securing frontier generative AI and released a paper on April 2. DeepMind focused on two of its four key risk areas: “misuse, misalignment, mistakes, and structural risks.” DeepMind is looking beyond current frontier AI to artificial general intelligence (AGI), human-level smarts, which could revolutionize healthcare and other industries or trigger technological chaos. There is some skepticism over whether AGI of that magnitude will ever exist. Asserting that…

Read More

Rethinking technology and IT's role in the era of agentic AI and digital labor

Posted on by Admin
Rethinking technology and IT's role in the era of agentic AI and digital labor

Cravetiger/Getty Images Generative AI, agentic AI, and other emerging technologies are morphing companies and driving businesses to rethink organizational structures and traditional roles. Also: Adobe brings four highly-requested Premiere Pro AI features out of beta That’s according to the latest report from consultant Accenture, ‘Rethinking IT operating models for the modern enterprise‘. The research suggests current IT processes will not allow businesses to stay ahead of technology disruption, so even though IT continues to play a…

Read More

The big VPN choice: System-wide or just in the browser? How to decide

Posted on by Admin
The big VPN choice: System-wide or just in the browser? How to decide

LagartoFilm/Getty Images If you use a VPN, are you using it all the time and for everything on your desktop or laptop?  You probably assume that more is always better, right? It is like Windows users thinking if they install two antivirus tools, they will be safer — not knowing that running two such tools could bring a system to a halt and wreak general havoc. Also: The best VPN services: Expert tested In this case,…

Read More

“생성형 AI가 이미 의료 산업을 변화시키고 있다” 맥킨지 보고서

Posted on by Admin
“생성형 AI가 이미 의료 산업을 변화시키고 있다” 맥킨지 보고서

한편 지금까지 생성형 AI 역량을 개발해 온 조직들은 대규모 구현을 통해 성공을 거두곤 했지만, 미래의 성공은 가치 중심 전략, 강력한 실행, 효과적인 관리에 달려 있다고 맥킨지는 밝혔다. 대부분의 응답자들은 조직이 생성형 AI 활용 사례를 구현했거나 구현 중이라고 답했으며, 개념 증명 작업조차 시작하지 않은 조직은 15%에 그쳤다. 이 밖의 주요 결과는 다음과 같다. 지불 조직, 의료 시스템, HST 그룹의 거의 절반(47%)이 이미 생성형 AI 도구를 도입했다. 의료 서비스 리더의 약 4분의 3은 이 기술이 행정 효율성과…

Read More

칼럼 | 현직 CISO들이 언급한 ‘취약점 관리 모범 사례’ 10가지

Posted on by Admin
칼럼 | 현직 CISO들이 언급한 ‘취약점 관리 모범 사례’ 10가지

한 CISO는 조직이 이를 한 단계 더 발전시켰다면서, 기업 인수 후 보안팀이 미리 준비한 프로그램을 통해 인수한 회사의 취약점 관리 프로그램을 기존 모델에 맞게 변환했으며, 진행 상황을 측정할 수 있는 지표까지 갖췄다고 설명했다. 4. 보안에 필요한 데이터 정의 명확히 하자면, 이는 처음부터 기술 목록을 작성하는 작업이 아니다. CISO들은 먼저 보유한 데이터를 평가하고 필요한 데이터와 비교 분석한다. 이 지식을 바탕으로 부족한 부분을 채울 수 있는 적절한 기술을 찾도록 팀원들에게 임무를 할당할 수 있다. 5. 취약점 관리에…

Read More
1 … 165 166 167 168 169 … 4,337