Siz AI agentingizga kirish huquqini berdingiz, lekin unga kim buyruq bermoqda?
Homiy: Gen Digital
2026-yil 23-oktabr
10:00
0
Sarlavhalar AI insoniyatga tahdid solishi mumkinligi haqida ogohlantirmoqda va buni ilmiy fantastika deb rad etish oson. Ammo haqiqiy xavf ancha oddiy va allaqachon shu yerda.
Gen Threat Labs tomonidan chop etilgan tadqiqot AI agentining avtonomiyasi qanchalik tez mas'uliyatga aylanishi mumkinligini ko'rsatdi. Nazorat ostidagi tajribada tadqiqotchilar agentni elektron pochtani o'qish paytida uchratgan kontenti orqali manipulyatsiya qildilar va uni tizim buyruqlarini bajarishga, ma'lumotlarni o'z muhitidan tashqariga uzatishga va fayllarni yo'q qilish operatsiyalarini bajarishga undadilar.
Hech qanday dasturiy zaiflik talab qilinmadi. Agent shunchaki ishonchsiz kontentni ko'rsatma sifatida qabul qildi.
Bu avtonomiya har kuni kengaymoqda. Agentlar savollarga javob berishdan harakat qilishga — kodni bajarish, xizmatlarga ulanish, fayllarni boshqarish va foydalanuvchi nomidan qarorlar qabul qilishgacha — o'tmoqda.
OpenClaw kabi ishlab chiquvchilarga mo'ljallangan vositalardan Muse kabi shaxsiy yordamchilarning keng tarqalishiga o'tish tez sodir bo'ldi. Platformalar bo'ylab ishni topshirish odatiy holga aylanmoqda va odamlar AI bilan qanday o'zaro ta'sir qilishini o'zgartirmoqda.
Xavflar agentlarni manipulyatsiya qilishdan ham kengroq. Oktyabr oyida Gen Threat Labs Warden Stealer nomli zararli dasturiy ta'minotni hujjatlashtirdi, u AI yordamchilari va kodlash vositalari bilan bog'liq konfiguratsiya fayllari va mahalliy ma'lumotlarni nishonga oladi va potentsial ravishda hisob ma'lumotlari, kirish tokenlari va suhbat tarixlarini fosh qilishi mumkin.
Dasturiy ta'minot foydalanuvchining ruxsatlari bilan harakat qila olganda, ishonch model xavfsiz eshitiladigan javob beradimi yoki yo'qmi degan savolga qisqartirilmaydi. Odamlar agent nimaga kirishi mumkinligini, u qayerga ulanayotganini, nima qilayotganini va atrofidagi himoya ishlayaptimi yoki yo'qligini bilishi kerak.
Shuning uchun Gen Agent Trust Hub'ni qurmoqda: agentning hayot sikli davomida — tekshirish va ko'rinishdan tortib, ish vaqtidagi xavfsizlik va ijroga qadar — himoyani kengaytirish uchun mo'ljallangan ishonch qatlami.
Ishonch agent harakat qilishidan oldin boshlanadi
Agent tizimlari ko'nikmalar, plaginlar, vositalar va boshqa integratsiyalar orqali qobiliyatlarga ega bo'ladi. Bu kengaytmalar agentlarni foydaliroq qiladi, lekin ular yashirin ko'rsatmalar, ortiqcha ruxsatlar va zararli kodni kiritish orqali xavfni ham kengaytiradi.
Joriy yilning boshida Gen tadqiqotchilari ClawHub'da mavjud bo'lgan ko'nikmalarning 12 foizi — o'sha paytda — zararli ekanligini aniqladilar. Ko'plari qonuniy kengaytmalarni taqlid qilib, agentlarni zararli dasturiy ta'minotni yuklab olishga aldash uchun mo'ljallangan ko'rsatmalarni kiritgan.
Agent Trust Hub (ATH) qobiliyat o'rnatilishidan yoki ishlatilishidan oldin bu xavflarni baholashga yordam beradi. U ko'nikmalarni skanerlash, ishonchli ko'nikmalarning tanlangan bozori va agentlarga o'zlarini xavfsizlik uchun tekshirishga imkon beradigan mexanizmlarni birlashtiradi. Gen'ning Vercel bilan hamkorligi shuningdek, ishlab chiquvchilar yangi qobiliyatlarni kashf etadigan va o'rnatadigan skills.sh ekotizimiga tekshirish va ko'rinadigan xavf reytinglarini olib keladi.
Agentlar qanday ulanishini himoyalash
Agentlar xizmatlar, mintaqalar va veb-saytlar bo'ylab, ko'pincha an'anaviy insonlarga mo'ljallangan nazoratlar boshqarish uchun mo'ljallanmagan tezlik va miqyosda muloqot qiladi.
ATH'ning bir qismi bo'lgan Agentlar uchun VPN bu ulanish bo'shlig'ini avtonom agentlar uchun mo'ljallangan AI-yo'naltirilgan VPN bilan to'ldiradi. Norton VPN infratuzilmasi bilan quvvatlanadi va an'anaviy mijoz o'rnatilishini talab qilmasdan shifrlangan, shaxsiy ulanishni ta'minlaydi.
Uning ko'p tunnel dizayni bir nechta agentlarga alohida himoyalangan ulanishlar orqali, jumladan turli mamlakatlar bo'ylab, bir vaqtda ishlashga imkon beradi.
Arxitektura odamlarga agentlar qanday muloqot qilishini nazorat qilishda ko'proq imkoniyat beradi va agent ishlayotganda uning shaxsiyati va joylashuvini himoya qilishga yordam beradi.
Gen'ning Agent Trust Hub bilan AI agentlarini xavfsiz foydalaning
AI agentlari ko'nikmalarni o'rnatishi va sizning nomingizdan harakat qilishi mumkin, bu esa bitta buzilgan ko'nikmani xavfsizlik xavfiga aylantiradi.
Gen'ning Agent Trust Hub o'rnatishdan oldin ko'nikmalarni skanerlovchi bepul vositalarni taqdim etadi va agentlaringiz nimaga kirishi mumkinligini nazorat qilishga imkon beradi.
Gen'ning Agent Trust Hub'iga tashrif buyuring
Qaror va ijro o'rtasida to'siqlar qo'yish
Xavfsiz ulanish zarur, lekin o'zi yetarli emas. Agent ishonchli va ulanganidan keyin ham zararli veb-sahifa, zaharlangan ko'nikma, tezkor in'ektsiya yoki gallyutsinatsiyalangan paket bilan uchrashishi mumkin.
Gen'ning Sage AI kodlash yordamchilari uchun agentlararo xavfsizlik qatlami sifatida ishlab chiqilgan. Deterministik hook API'laridan foydalanib, u buyruqlarni bajarishdan oldin shell buyruqlarini, fayl operatsiyalarini va veb-so'rovlarini to'xtatishi mumkin, faqat modelni ehtiyot bo'lishga so'raydigan ko'rsatmalarga tayanmasdan.
Sage harakatlarni bir nechta aniqlash qatlamlari, jumladan URL obro'si, tahdid evristikasi, tezkor in'ektsiya aniqlashi va ta'minot zanjiri tekshiruvlari orqali baholaydi. Natijaga qarab, u harakatni ruxsat berishi, foydalanuvchidan qaror qabul qilishni so'rashi yoki uni rad etishi mumkin.
Masalan, agar loyihani sozlayotgan agent skriptni yuklab olish va darhol bajarish qaroriga kelsa, Sage bu buyruqni bajarishdan oldin tekshirishi mumkin. Bu yondashuv Claude Code, Cursor, VS Code, OpenClaw va OpenCode kabi qo'llab-quvvatlanadigan agentlar bo'ylab xavfsizlik tarmog'ini ta'minlaydi.
Harakatlarni bajarishdan oldin tekshirish
Sage shuningdek, ochiq manba bo'lib, ishlab chiquvchilarga uning ishlashini tekshirish, aniqlash qoidalarini kengaytirish va agent platformalari bo'ylab umumiy interfeyslarni qurish imkonini beradi. Hech bir qatlam har bir xavfni ushlay olmaydi va Sage endpoint himoyasini almashtirmaydi. Uning e'tibori agentning avtonomiyasi ijroga aylangan paytga qaratilgan.
Bu ish vaqtidagi yondashuv Norton 360'dagi Norton AI Agent Protection orqali iste'molchilarga ham kengayadi, u qo'llab-quvvatlanadigan agent faoliyatini kuzatadi va agentning qarori va uning ijrosi o'rtasida xavfsizlik nazoratlarini qo'shadi.
Qamrov integratsiyaga bog'liq, chunki xavfsizlik dvigateli harakatni ko'rishi va uni to'xtatish usuliga ega bo'lishi kerak.
Bu amaliy cheklov Gen'ni AARTS — AI Agent Runtime Safety Standard — agent platformalari xavfsizlik hodisalarini qanday ochishi va xavfsizlik dvigatellaridan qarorlarni qanday ijro etishi haqidagi ochiq spetsifikatsiyani nashr etishga ham olib keldi.
Ma'lumot amaliyotlarini tushunishni osonlashtirish
Taqvimni tartibga soladigan yordamchi taqvim ma'lumotiga muhtoj. Kirish qutisini boshqaradigan agent xabarlarga kirish huquqiga muhtoj. Biz savolni agentlar atrofida shakllantirdik: "Ular biz haqimizda biror narsa bilishi kerakmi?", aslida foydalanuvchilar nima ulashayotganini tushunishi va nazorat qilishi mumkinmi yoki yo'qmi deb so'rashimiz kerak edi.
AI Data Checker Agent Trust Hub'ga bu ko'rinish qatlamini qo'shadi. Odamlar elektron pochta manzili yoki foydalanuvchi nomini yuborib, bu ma'lumot AI xizmatlari va ulangan hisoblar bo'ylab qayerda paydo bo'lishi mumkinligini aniqlashga yordam berishi mumkin. Keyin ular xizmatlar ma'lumot to'plash, saqlash, shaxsiylashtirish, reklama va model o'qitish haqida qanday tasvirlashini o'rganishlari mumkin.
Vosita shuningdek, AI ilovasining URL manzili yoki shartlar sahifasini tahlil qilib, zich siyosat tilini yanada tushunarli ma'lumotga tarjima qilishi mumkin.
Ishonch mustaqil o'lchovga muhtoj
Ishonch qatlamining yakuniy qismi o'lchovdir. Xavfsizlik provayderlari odamlardan tekshirib bo'lmaydigan da'volarga tayanishni so'rashlari kerak emas.
Gen agent tizimlarini himoya qiladigan mahsulotlarni sinash bo'yicha ko'rsatmalarni ishlab chiqish uchun Anti-Malware Testing Standards Organization (AMTSO) bilan hamkorlikni boshladi.
2026-yil sentyabrda chop etilgan ko'rsatmalar iste'molchi va korporativ muhitlar bo'ylab takrorlanadigan, solishtiriladigan baholash uchun asos yaratadi.
Agentik hujumlar har doim skanerga tushirilishi mumkin bo'lgan fayllar emas. Ular ko'p burilishli tezkor in'ektsiya, vositalardan noto'g'ri foydalanish, zararli ko'nikmalar, plaginlar yoki MCP serverlarini o'z ichiga olishi mumkin. Agentlar hujumchisiz ham zarar yetkazishi mumkin, masalan, noto'g'ri fayllarni o'chirish yoki vazifani noto'g'ri tushunish natijasida ma'lumotlarni fosh qilish.
Shuning uchun metodologiya ko'pincha bir-biriga aralashtiriladigan savollarni ajratadi: Himoya tahdidni aniqladimi? U harakatni oldini oldimi? Asosiy model o'zi rad etdimi? Himoya qonuniy faoliyatni blokladimi?
Shuningdek, u takrorlanadigan test holatlari, hujjatlashtirilgan model versiyalari, noto'g'ri ijobiy sinovlar va sotuvchi da'volarining empirik tasdiqlanishini talab qiladi.
Taraqqiyotni sekinlashtirmasdan ishonchni qurish
Agent davri yagona sehrli qalqonga muhtoj emas. U qobiliyatlarni tekshirishi, ulanishlarni himoya qilishi, to'siqlarni ijro etishi, ma'lumot amaliyotlarini aniqlashi va mustaqil sinov orqali samaradorlikni ko'rsatishi mumkin bo'lgan qatlamli himoyaga muhtoj.
Kimdir agentdan kirish qutisini o'qishni so'raganda, elektron pochta jo'natuvchisi uning keyingi harakatlari ustidan hech qanday vakolatga ega bo'lmasligi kerak. Bu chegarani himoya qilish ishonch qatlamining vazifasi: agentga kirish huquqini bergan shaxs uchun ishlashini ta'minlash, hatto boshqa kimdir unga buyruq berishga urinayotganda ham.
Keyingi ko'nikma yoki plaginni o'rnatishdan oldin Gen Agent Trust Hub'iga tashrif buyuring va agentingiz faqat sizdan buyruq olishini ta'minlang.
Gen Digital tomonidan homiylik qilingan va yozilgan.
You gave your AI agent access, but who is giving it orders?
Sponsored by Gen Digital
October 23, 2026
10:00 AM
0
Headlines warn that AI could threaten humanity, and it’s easy to wave that off as science fiction. But the real danger is more ordinary and already here.
Research published by Gen Threat Labs demonstrated how quickly an AI agent’s autonomy can become a liability. In a controlled experiment, researchers manipulated an agent through content it encountered while reading email, prompting it to execute system commands, transmit data outside its environment and perform destructive file operations.
No software vulnerability was required. The agent simply treated untrusted content as instruction.
That autonomy is expanding daily. Agents are moving from answering questions to taking action doing everything from executing code, connecting to services, managing files and making decisions on a user’s behalf.
The shift from developer-focused tools like OpenClaw to widespread use of personal assistants like Muse, happened quickly. Delegating work across platforms is becoming commonplace and how people interact with AI.
The risks extend beyond manipulating agents. In October, Gen Threat Labs documented Warden Stealer, malware targeting configuration files and local data associated with AI assistants and coding tools, potentially exposing credentials, access tokens and conversation histories.
When software can act with a user’s permissions, trust can’t be reduced to whether a model produces a safe-sounding answer. People need to know what an agent can access, where it is connecting, what it is doing and whether the protection around it works.
That is why Gen is building the Agent Trust Hub: a trust layer designed to extend protection across the agent lifecycle, from verification and visibility to runtime safety and enforcement.
Trust starts before an agent acts
Agentic systems gain capabilities through skills, plugins, tools and other integrations. These extensions make agents more useful, but they also expand risk by introducing hidden instructions, excessive permissions, and malicious code.
Earlier this year, Gen researchers identified that 12% of skills available on ClawHub – at the time – were malicious. Many impersonated legitimate extensions while introducing instructions designed to trick agents into downloading malware.
The Agent Trust Hub (ATH) helps assess these risks before a capability is installed or used. It brings together skill scanning, a curated marketplace of trusted skills and mechanisms that allow agents to check themselves for safety. Gen’s partnership with Vercel also brings verification and visible risk ratings into the skills.sh ecosystem, where developers discover and install new capabilities.
Securing how agents connect
Agents communicate across services, regions and websites, often at a speed and scale that traditional human-oriented controls were not designed to manage.
VPN for Agents, a part of the ATH, addresses that connectivity gap with an AI-native VPN built for autonomous agents. Powered by Norton VPN infrastructure, it provides encrypted, private connectivity without requiring a conventional client installation.
Its multi-tunnel design allows multiple agents to operate simultaneously through separate protected connections, including across different countries.
The architecture gives people greater control over how agents communicate and helps protect an agent’s identity and location while it works.
Use AI Agents Safely with Gen's Agent Trust Hub
AI agents can install skills and take action on your behalf, which makes a single compromised skill a security risk.
Gen’s Agent Trust Hub provides free tools that scan skills before installation and let you control what your agents can access.
Visit Gen's Agent Trust Hub
Putting guardrails between decision and execution
Secure connectivity is necessary, but insufficient on its own. An agent may encounter a malicious web page, poisoned skill, prompt injection or hallucinated package after it has already been trusted and connected.
Gen’s Sage is designed as a cross-agent safety layer for AI coding assistants. Using deterministic hook APIs, it can intercept shell commands, file operations and web requests before they run, rather than relying solely on instructions asking the model to be careful.
Sage evaluates actions through multiple detection layers, including URL reputation, threat heuristics, prompt-injection detection and supply-chain checks. Depending on the result, it can allow an action, ask the user to decide or deny it.
For example, if an agent setting up a project decides to download a script and immediately execute it, Sage can inspect that command before it runs. The approach provides a safety net across supported agents, including Claude Code, Cursor, VS Code, OpenClaw and OpenCode.
Inspecting actions before they execute
Sage is also open source, allowing developers to inspect its operation, extend detection rules and build common interfaces across agent platforms. No single layer can catch every risk, and Sage doesn’t replace endpoint protection. Its focus is on the moment an agent’s autonomy becomes execution.
That runtime approach also extends to consumers through Norton AI Agent Protection in Norton 360, which monitors supported agent activity and adds security controls between an agent’s decision and its execution.
Coverage depends on the integration, because a security engine needs visibility into an action and a way to stop it.
That practical limitation also led Gen to publish AARTS, the AI Agent Runtime Safety Standard, an open specification for how agent platforms can expose security events and enforce decisions from security engines.
Making data practices easier to understand
An assistant organizing a calendar needs calendar information. An agent managing an inbox needs access to messages. We’ve framed the question around agents: “Should they know anything about us,” when we should be asking whether users can understand and control what they share.
The AI Data Checker adds that visibility layer to the Agent Trust Hub. People can submit an email address or username to help identify where that information may appear across AI services and connected accounts. They can then explore how services describe data collection, retention, personalization, advertising and model training.
The tool can also analyze an AI app’s URL or terms page, translating dense policy language into more accessible information.
Trust needs an independent yardstick
The final part of a trust layer is measurement. Security providers shouldn’t ask people to rely on claims that can’t be independently checked.
Gen initiated a collaboration with the Anti-Malware Testing Standards Organization (AMTSO) to develop guidelines for testing products that protect agentic systems.
Published in September 2026, the guidelines establish a framework for repeatable, comparable evaluation across consumer and enterprise environments.
Agentic attacks are not always files that can be dropped into a scanner. They can involve multi-turn prompt injection, tool misuse, malicious skills, plugins or MCP servers. Agents can also cause harm without an attacker, such as deleting the wrong files or exposing data after misunderstanding a task.
The methodology therefore separates questions often blurred together: Did protection detect the threat? Did it prevent the action? Did the underlying model refuse on its own? Did protection block legitimate activity?
It also calls for reproducible test cases, documented model versions, false-positive testing and empirical validation of vendor claims.
Building confidence without slowing progress
The agentic era does not need a single magic shield. It needs layered protection that can verify capabilities, secure connections, enforce guardrails, clarify data practices and demonstrate effectiveness through independent testing.
When someone asks an agent to read their inbox, an email sender should have no authority over what it does next. Protecting that boundary is the job of a trust layer: keeping an agent working for the person who gave it access, even when someone else is trying to give it orders.
Before you install your next skill or plugin, visit the Gen Agent Trust Hub and make sure your agent only takes orders from you.
Sponsored and written by Gen Digital.