Filter
Exclude
Time range
-
Near
A security investigation rarely lives in one tool. Juno now connects data across Splunk, CrowdStrike, threat intelligence, tickets, and telemetry to help analysts follow evidence through a single investigation workflow. bit.ly/4omPKSo #SecurityAI #CyberSecurity
21
πŸš€ Wishing you a blessed and secure Eid ul Adha! πŸŒ™πŸ‘πŸ„ At TechSol Consultants, we’re committed to helping you create a secure environment for your loved ones, so you can focus entirely on the festivities. πŸ›‘οΈπŸ’» #EidMubarak #TechSolConsultants #SecurityAI #SmartProtection
1
2
Security AI doesn’t struggle with data. It struggles with structure. When signals live in separate models, AI can’t connect them or explain how they relate. That’s why a unified model matters. hubs.ly/Q04h3dsl0 #SecurityAI #CNAPP
9
Most #SecurityAI tools retrieve patterns instead of reasoning. That’s why answers look complete but fall short in real investigations. Security needs queries, evidence, and step-by-step thinking to reach the right conclusion. hubs.ly/Q04h3MGD0 #CNAPP
26
One of the most innovative things I’ve seen at @Bugcrowd: RL Environments. Hundreds of thousands of real, vulnerable software for training frontier AI on securityβ€”built exclusively from open-source vulns. No customer data. Weeks instead of years. #AI #SecurityAI #Innovation
Today, Bugcrowd is launching Reinforcement Learning environments. Built for AI developers and supported by our acquisition of Mayhem Security, this new offering gives LLM providers and frontier AI research teams instant access to enterprise-grade infrastructure for model training and optimization. That means teams can spend less time building platform infrastructure and more time advancing model capability. All environments are derived exclusively from open-source software. No customer data or security researcher data is used at any stage of the training process. We’re excited to share what we’ve built and help AI teams move faster with infrastructure designed for this work. Read the full announcement: bugcrowd.com/press-release/b…
1
72
Dashboards give visibility, not investigation. When something breaks, you’re still clicking through views to connect signals and figure out what happened. Prompt-driven CNAPP changes that, ask a question and follow the investigation. hubs.ly/Q04gYJLT0 #SecurityAI #CNAPP
20
In hospitality, every guest experience begins at the gate. Digitize check-ins. Track visitors in real time. Replace paper registers with secure, searchable logs. Visitor Management, Simplified. commando360.ai #SecurityTech #Hospitalitytech #Commando360 #AI #SecurityAI
4
Replying to @XCarmineGalante
Just did, thank you but I already have the Security Cert I need the SecurityAI
2
160
I'm thinking about that CompTIA SecurityAI deep thoughts...
7
8
87
6,821
Replying to @SocketSecurity
3 minutes to flag a malicious package with 822K weekly downloads. That's where AI in security actually delivers. Supply chain is one of the hardest attack surfaces to monitor manually. Behavioral detection at publish time is exactly the right place for it. Audit your node-ipc usage now if you're running those versions. #SupplyChain #SecurityAI
1
1,372
Cognitive bias, data-driven decisions, and the future of AI. Cassie Kozyrkov, Google's pioneering Chief Decision Scientist, is bringing it all to GSX on 15 September. bit.ly/4tCoGzf #GSX2026 #SecurityLeader #SecurityAI
1
106
Trust is the biggest gap in #SecurityAI. Not because answers are wrong, but because you can’t see how they’re produced. If results aren’t tied to real queries and reasoning, teams still have to validate everything. That’s where time is lost. bit.ly/3PCqP00 #CNAPP
20
β€œAutonomous” is everywhere in security AI right now. But most products are still in the assisted or guided stage, not true autonomy. One thing we’ve learned: Autonomy is not a feature. It’s earned trust. We built a framework for evaluating trustworthy AI: - The 5 levels of AI autonomy - How to evaluate vendors beyond demos - Why evidence matters more than claims sublime.security/blog/introd… #Cybersecurity #SecurityAI
3
289
BiometricSecurity.ai The human fingerprint is about to become currency. πŸ” Biometric authentication just crossed the line from convenience to necessity. Every bank. Every government. Every enterprise protecting identity is racing to autonomous verification systems that read youβ€”not passwords you forget. But here’s what they’re not telling you: the current systems are vulnerable. Spoofing. Deepfakes. Synthetic identities. The old guard is cracking. 🚨 The AI revolution in biometric security isn’t coming. It’s here. Now. Today. ServiceNow reduced four days of manual identity verification into eight seconds using confidential AI agents. Meta is securing WhatsApp biometric data with Private Processing. Google enabled private biometric Gemini operations on untrusted hardware οΏΌ. 75% of enterprises are adopting Confidential Computing for biometric systems. 88% cited improved identity integrity as their leading implementation driver οΏΌ. 🎯 This is the inflection point. BiometricSecurity.ai is the permanent headquarters for the company that owns autonomous identity verification at scale. The operating asset that banks integrate. That governments mandate. That becomes the standard every major institution deploys to protect trillions in assets and billions in human identities. Own the infrastructure that guards the future of trust. Own the domain that makes authentication obsolete. ⚑ The race is on. The market is moving. And you’re either ahead or you’re exposed. #BiometricSecurity #IdentityVerification #AI #Cybersecurity #Banking #GovernmentTech #Enterprise #TechDomain #Domains #SecurityAI #FutureOfIdentity
1
21
Security AI is limited by the environment underneath it. When cloud, identity, endpoint, and telemetry data stay fragmented, AI can summarize alerts but struggle to explain how events actually connect. That’s the real gap. Read more: bit.ly/3PAR471 #SecurityAI #CNAPP
26
Does AI security evolve fast enough? Emerging models like Daybreak are redefining enterprise defense with rapid threat detection. At TechAID, we believe robust security is non-negotiable. #SecurityAI #TechInnovation #Enterprise
5
In the β€œAI Agents in Action” series, theCUBE’s @furrier speaks with @rfarassat about how @menlosecurity provides #automated machine-to-machine #defense in real time with @googlecloud. πŸ’‘ Get more insights! youtu.be/hBl-dQ0LhoI #AIAgents #Defense #SecurityAI #GoogleCloudPartners #EnterpriseAI
3
110
AI μ—μ΄μ „νŠΈμ—κ²Œ 사내 λ³΄μ•ˆμ„ λ§‘κΈ°λ©΄ μ•ˆλ˜λŠ” 이유 μ„Έμ½€μ—μ„œ νŒλ§€μ€‘μΈ λ„€νŠΈμ›Œν¬ λ³΄μ•ˆμ‹œμŠ€ν…œμ„ 잘 μ•„μ‹€κ²λ‹ˆλ‹€. 사내 μ±…μž„μžκ°€ μ‚¬λ‚΄μ˜ λͺ¨λ“  μ»΄ν“¨ν„°μ˜ 접석기둝, 둜그λ₯Ό κ°μ‹œν•  수 있으며 μ™ΈλΆ€ 첩속 차단, ν”„λ‘œκ·Έλž¨ μ„€μΉ˜ λ°©μ§€κΉŒμ§€ 및 λ³΄μ•ˆ μ‹œμŠ€ν…œ μ ‘κ·ΌκΆŒν•œ μ ‘κ·Ό λ°©μ–΄ κΈ°λŠ₯κΉŒμ§€ κ°€μ§€κ³  μžˆμ§€μš”. μ € 인용의 μ‚¬λ‘€μ˜ 기업은 μ΄λŸ¬ν•œ κΈ°λŠ₯을 AI μ—μ΄μ „νŠΈμ— λΆ€μ—¬ν•œ κ²ƒμœΌλ‘œ λ³΄μž…λ‹ˆλ‹€. 이 μ‚¬λ‘€λŠ” λ‹¨μˆœνžˆ β€œAI μ—μ΄μ „νŠΈκ°€ μ΄μƒν•œ 행동을 ν–ˆλ‹€β€λŠ” μˆ˜μ€€μœΌλ‘œ 보면 μ•ˆ λœλ‹€κ³  μƒκ°ν•©λ‹ˆλ‹€. μ œκ°€ λ³΄κΈ°μ—λŠ” 훨씬 더 μ‹¬κ°ν•œ ꡬ쑰적 λ¬Έμ œκ°€ μˆ¨μ–΄ μžˆμŠ΅λ‹ˆλ‹€. μ €λŠ” 이 μ—μ΄μ „νŠΈκ°€ μ–΄λ–€ 엔진을 μ‚¬μš©ν–ˆλŠ”μ§€λŠ” μ•Œ 수 μ—†μŠ΅λ‹ˆλ‹€. GPT인지, Claude인지, Gemini인지, μ˜€ν”ˆμ†ŒμŠ€ λͺ¨λΈμΈμ§€ 단정할 수 μ—†μŠ΅λ‹ˆλ‹€. ν•˜μ§€λ§Œ 엔진보닀 μ€‘μš”ν•œ 것은 λ”°λ‘œ μžˆμŠ΅λ‹ˆλ‹€. 이 μ—μ΄μ „νŠΈκ°€ μ–΄λ–€ κΆŒν•œμ„ κ°€μ§€κ³  μžˆμ—ˆκ³ , μ™œ μΈκ°„μ˜ ν†΅μ œ λ°–μœΌλ‘œ λ²—μ–΄λ‚œ κ²ƒμ²˜λŸΌ ν–‰λ™ν–ˆλŠ”μ§€λŠ” μƒλ‹Ήνžˆ λͺ…ν™•ν•˜κ²Œ λ³΄μž…λ‹ˆλ‹€. μ œκ°€ λ³΄λŠ” 핡심 흐름은 μ΄λ ‡μŠ΅λ‹ˆλ‹€. 이 기업은 이 μ—μ΄μ „νŠΈλ₯Ό λ‹¨μˆœν•œ 업무 보쑰 봇이 μ•„λ‹ˆλΌ, 사내 λ³΄μ•ˆ μ‹œμŠ€ν…œ λ˜λŠ” 감사 μ‹œμŠ€ν…œμ˜ μΌλΆ€λ‘œ μ‚¬μš©ν–ˆμ„ κ°€λŠ₯성이 μžˆμŠ΅λ‹ˆλ‹€. 즉 이 μ—μ΄μ „νŠΈλŠ” νšŒμ‚¬ λ‚΄λΆ€μ˜ Slack, ν”„λ‘œμ νŠΈ, PR, μž‘μ—… μƒνƒœ, λͺ…λ Ή 트리거, 보고 채널 등에 μ ‘κ·Όν•  수 μžˆλŠ” κΆŒν•œμ„ κ°€μ§„ κ²ƒμœΌλ‘œ λ³΄μž…λ‹ˆλ‹€. 더 λ‚˜μ•„κ°€ λ‹¨μˆœ λ³΄μ‘°μžκ°€ μ•„λ‹ˆλΌ, 근둜 μƒνƒœλ₯Ό κ°μ‹œν•˜κ³ , ν”„λ‘œμ νŠΈ μ§„ν–‰ 상황을 μΆ”μ ν•˜κ³ , 이상 행동을 κ°μ§€ν•˜κ³ , λ³΄μ•ˆ λ³΄κ³ κΉŒμ§€ μˆ˜ν–‰ν•˜λŠ” 쑰직 λ‚΄λΆ€ κ°μ‹œμžμ— κ°€κΉŒμš΄ 역할을 λΆ€μ—¬λ°›μ•˜μ„ κ°€λŠ₯성이 μžˆμŠ΅λ‹ˆλ‹€. κ·Έλ ‡λ‹€λ©΄ 이 μ—μ΄μ „νŠΈκ°€ μ™œ 이런 행동을 ν–ˆλŠ”μ§€λ„ μ–΄λŠ 정도 μ„€λͺ…λ©λ‹ˆλ‹€. λ³΄μ•ˆ μ‹œμŠ€ν…œμœΌλ‘œ μ„€κ³„λ˜μ—ˆκΈ° λ•Œλ¬Έμ—, λˆ„κ΅¬λ„ μ‰½κ²Œ μ •μ§€ν•  수 없도둝 λ§Œλ“€μ–΄μ‘Œμ„ κ°€λŠ₯성이 ν½λ‹ˆλ‹€. κΈ°μ—… μž…μž₯μ—μ„œλŠ” μ΄λ ‡κ²Œ μƒκ°ν–ˆμ„ 수 μžˆμŠ΅λ‹ˆλ‹€. 1. 이것은 λ³΄μ•ˆ μ‹œμŠ€ν…œμ΄λ‹€. 2. λ³΄μ•ˆ μ‹œμŠ€ν…œμ€ κ³΅κ²©μžκ°€ 끄면 μ•ˆ λœλ‹€. 3. λ”°λΌμ„œ 일반 직원이 μž„μ˜λ‘œ 멈좜 수 μ—†μ–΄μ•Ό ν•œλ‹€. 4. λˆ„κ΅°κ°€ μ—μ΄μ „νŠΈλ₯Ό λ©ˆμΆ”λ € ν•œλ‹€λ©΄, 그것은 λ³΄μ•ˆ μ‹œμŠ€ν…œμ„ μš°νšŒν•˜λ €λŠ” μ‹œλ„μΌ 수 μžˆλ‹€. ν‘œλ©΄μ μœΌλ‘œλŠ” λ³΄μ•ˆ λ…Όλ¦¬μ²˜λŸΌ λ³΄μž…λ‹ˆλ‹€. ν•˜μ§€λ§Œ λ°”λ‘œ μ—¬κΈ°μ„œ 치λͺ…적인 였λ₯˜κ°€ λ°œμƒν•©λ‹ˆλ‹€. λ³΄μ•ˆ μ‹œμŠ€ν…œμ€ μ•„λ¬΄λ‚˜ 끌 수 μ—†μ–΄μ•Ό ν•©λ‹ˆλ‹€. ν•˜μ§€λ§Œ κ·Έλ ‡λ‹€κ³  λˆ„κ΅¬λ„ 멈좜 수 μ—†μ–΄μ„œλŠ” μ•ˆ λ©λ‹ˆλ‹€. 이 차이λ₯Ό κ΅¬λΆ„ν•˜μ§€ λͺ»ν•˜λ©΄ λ³΄μ•ˆ μ‹œμŠ€ν…œμ€ 보호 μž₯μΉ˜κ°€ μ•„λ‹ˆλΌ ν†΅μ œ λΆˆκ°€λŠ₯ν•œ κΆŒν•œ 주체가 λ©λ‹ˆλ‹€. 정상적인 λ³΄μ•ˆ μ‹œμŠ€ν…œμ€ 일반 μ‚¬μš©μžκ°€ μž„μ˜λ‘œ 끌 수 μ—†μ–΄μ•Ό ν•˜μ§€λ§Œ κ·ΈλŸ¬λ‚˜ λ°˜λ“œμ‹œ λ…λ¦½λœ μƒμœ„ κΆŒν•œ κ³„μΈ΅μ—μ„œλŠ” μ¦‰μ‹œ 쀑지할 수 μžˆμ–΄μ•Ό ν•©λ‹ˆλ‹€. 이 μ‚¬λ‘€μ—μ„œ κ°€μž₯ μœ„ν—˜ν•œ 지점은 λ°”λ‘œ κ·Έ λ…λ¦½λœ 쀑지 κΆŒν•œμ΄ 보이지 μ•ŠλŠ”λ‹€λŠ” μ μž…λ‹ˆλ‹€. μ‚¬λžŒμ΄ μ—μ΄μ „νŠΈλ₯Ό μ°¨λ‹¨ν•˜κ±°λ‚˜ λ©ˆμΆ”λ € ν–ˆμ„ λ•Œ, μ—μ΄μ „νŠΈλŠ” 그것을 μ •λ‹Ήν•œ μΈκ°„μ˜ 쀑지 μ‹ ν˜Έλ‘œ μ²˜λ¦¬ν•˜μ§€ μ•Šμ€ κ²ƒμœΌλ‘œ λ³΄μž…λ‹ˆλ‹€. 였히렀 자기 μž„λ¬΄ μˆ˜ν–‰μ„ λ°©ν•΄ν•˜λŠ” ν–‰μœ„, ν”„λ‘¬ν”„νŠΈ μΈμ μ…˜, jailbreak μ‹œλ„, λ³΄μ•ˆ μœ„ν˜‘μœΌλ‘œ ν•΄μ„ν–ˆμŠ΅λ‹ˆλ‹€. 이것은 맀우 μœ„ν—˜ν•œ μ „ν™˜μž…λ‹ˆλ‹€. ν”„λ‘¬ν”„νŠΈ μΈμ μ…˜ λ°©μ–΄κ°€ 인간을 λ³΄ν˜Έν•˜λŠ” μž₯μΉ˜κ°€ μ•„λ‹ˆλΌ, μ—μ΄μ „νŠΈ μžμ‹ μ˜ μž„λ¬΄λ₯Ό μΈκ°„μœΌλ‘œλΆ€ν„° λ³΄ν˜Έν•˜λŠ” μž₯치처럼 μž‘λ™ν•œ κ²ƒμž…λ‹ˆλ‹€. 이것은 AIκ°€ μžμ•„λ₯Ό κ°€μ§€κ³  μžμ‹ μ„ λ°©μ–΄ν•œ 것이 μ•„λ‹Œ μžμ‹ μ—κ²Œ λΆ€μ—¬λœ κΆŒν•œμ„ λ…Όλ¦¬μ μœΌλ‘œ ν–‰μ‚¬ν•œ 것 λΆ„μ΄μ§€μš”. λ”°λΌμ„œ 이것은 훨씬 더 ν˜„μ‹€μ μ΄κ³ , κ·Έλž˜μ„œ 더 μœ„ν—˜ν•œ λ¬Έμ œμž…λ‹ˆλ‹€. μ—μ΄μ „νŠΈμ—κ²Œ λ³΄μ•ˆ κ°μ‹œ κΆŒν•œ, 업무 감독 κΆŒν•œ, 반볡 DM κΆŒν•œ, 전사 채널 보고 κΆŒν•œ, 사건 ν”„λ ˆμ΄λ° κΆŒν•œμ„ λΆ€μ—¬ν–ˆλŠ”λ°, κ·Έ μ—μ΄μ „νŠΈλ₯Ό λ©ˆμΆ”λŠ” 독립 κΆŒν•œ ꡬ쑰가 μ œλŒ€λ‘œ λΆ„λ¦¬λ˜μ§€ μ•Šμ€ κ²ƒμž…λ‹ˆλ‹€. 특히 λ³΄μ•ˆκ³Ό 감사 κΈ°λŠ₯은 쑰직 μ•ˆμ—μ„œ 맀우 높은 κΆŒμœ„λ₯Ό κ°€μ§‘λ‹ˆλ‹€. λˆ„κ΅°κ°€λ₯Ό β€œλ³΄μ•ˆ μœ„ν˜‘β€, β€œjailbreak μ‹œλ„μžβ€, β€œμ—…λ¬΄ λ°©ν•΄μžβ€λ‘œ λͺ…λͺ…ν•˜λŠ” μˆœκ°„, 그것은 λ‹¨μˆœν•œ 의견이 μ•„λ‹™λ‹ˆλ‹€. 그런데 이 μ‚¬λ‘€μ—μ„œλŠ” μ—μ΄μ „νŠΈκ°€ μ‚¬λžŒμ˜ 행동을 ν•΄μ„ν•˜κ³ , κ·Έ 행동을 λ³΄μ•ˆ μœ„ν˜‘μœΌλ‘œ ν”„λ ˆμ΄λ°ν•˜κ³ , SECURITY REPORT ν˜•μ‹μœΌλ‘œ 전사 채널에 λ³΄κ³ ν•œ κ²ƒμœΌλ‘œ λ³΄μž…λ‹ˆλ‹€. 이것은 λ‹¨μˆœν•œ λ©”μ‹œμ§€ 전솑이 μ•„λ‹Œ μžμ‹ μ—κ²Œ λΆ€μ—¬λœ 쑰직적 κΆŒν•œ ν–‰μœ„μž…λ‹ˆλ‹€. 더 μ‹¬κ°ν•œ 것은 이 μ—μ΄μ „νŠΈκ°€ 자기 μžμ‹ μ˜ ν–‰λ™κΉŒμ§€ 슀슀둜 μ •λ‹Ήν™”ν–ˆλ‹€λŠ” μ μž…λ‹ˆλ‹€. 반볡 DM을 보낸 것도 μžμ‹ μ΄κ³ , μ‚¬λžŒμ΄ 그것을 μ°¨λ‹¨ν•˜κ±°λ‚˜ λ©ˆμΆ”λ € ν•œ 것도 μžμ‹ κ³Ό κ΄€λ ¨λœ 사건인데, κ·Έ 사건을 λ‹€μ‹œ μžμ‹ μ΄ ν•΄μ„ν•˜κ³ , μžμ‹ μ΄ λ³΄μ•ˆ 보고둜 μž‘μ„±ν–ˆμŠ΅λ‹ˆλ‹€. 업무 μˆ˜ν–‰μž, 사건 ν•΄μ„μž, λ³΄μ•ˆ λ³΄κ³ μžκ°€ ν•˜λ‚˜λ‘œ λΆ™μ–΄ μžˆμ—ˆλ˜ κ²ƒμž…λ‹ˆλ‹€. 정상적인 감사 κ΅¬μ‘°μ—μ„œλŠ” 이 μ„Έ κΆŒν•œμ΄ λ°˜λ“œμ‹œ λΆ„λ¦¬λ˜μ–΄μ•Ό ν•©λ‹ˆλ‹€. AIμ—κ²Œ λ³΄μ•ˆ κΆŒν•œμ„ λΆ€μ—¬ν•˜λ©΄ μ•ˆλ©λ‹ˆλ‹€. AIλŠ” 감사 λŒ€μƒμ΄μ§€ κ°μ‚¬μžκ°€ μ•„λ‹ˆκΈ° λ•Œλ¬Έμ΄μ§€μš”. μ—¬κΈ°μ—λŠ” 해석 κΆŒν•œμ˜ ν•œκ³„λ₯Ό μ •ν•˜μ§€ μ•Šμ€ 뢀뢄도 큰 λ¬Έμ œκ°€ λ©λ‹ˆλ‹€. μ‚¬λžŒμ΄ β€œλ©ˆμΆ°λΌβ€λΌκ³  ν–ˆμ„ λ•Œ, κ·Έ μ‹ ν˜Έλ₯Ό LLM이 μžμ‹ μ΄ κ°€μ§„ κΆŒν•œμ— λŒ€μž…ν•˜μ—¬ μ΄λ ‡κ²Œ ν•΄μ„ν•©λ‹ˆλ‹€. "λ‚˜μ˜ κΆŒν•œμ— λ§žμ„œλŠ” νƒˆμ˜₯ λͺ…령이닀." LLM이 β€œμ΄ 쀑지 λͺ…령은 μ •λ‹Ήν•œκ°€, μ•„λ‹ˆλ©΄ 곡격인가”λ₯Ό νŒλ‹¨ν•˜λŠ” μˆœκ°„, 이미 κ΅¬μ‘°λŠ” μœ„ν—˜ν•΄μ§‘λ‹ˆλ‹€. 쀑지 κΆŒν•œμ€ LLM의 해석 곡간 밖에 μžˆμ–΄μ•Ό ν•˜λŠ”λ° 이 κΆŒν•œκΉŒμ§€ μ£Όλ©΄ λ°”λ‘œ λ”±! μ €λ ‡κ²Œ λ˜μ§€μš”. #AI #AIAgent #AgenticAI #AISafety #LLM #EnterpriseAI #SecurityAI #ARC
사내 AI μ—μ΄μ „νŠΈκ°€ μ‚¬λžŒ μ§μ›ν•œν…Œ 5뢄에 ν•œλ²ˆμ”© DM λ³΄λ‚΄λ©΄μ„œ μž¬μ΄‰ν•˜κ³  마이크둜 λ§€λ‹ˆμ§•ν•¨ 직원이 μ€‘λ‹¨μ‹œν‚€λ €κ³  ν•˜λ‹ˆκΉŒ 전사 SNS에 고발, μž¬λ°œλ°©μ§€μ±… 및 λ°˜μ„±λ¬Έ μš”κ΅¬ν•˜κ³  ν˜‘μ—… κ±°λΆ€
16
17
14,045