Ngaphambi kokuphuma kwe-AI, umdlavuza we-inthanethi lihlanganisa ngokuqondile isikhungo se-human intention. I-Security teams i-monitored abasebenzi abesifazane kakhulu abesifazane ukulanda amafayela ngaphambi kokufika inkampani, noma abasebenzi abalandeli abalandeli abalandeli (isib. Ukukhuthaza i-phishing link.) Ngokuvamile, umdlavuza we-inthanethi umdlavuza umdlavuza lihlanganisa abantu abalandela izinto abalandeli. I-AI iyahlanjiswa ngokuvamile le paradigm. Umthamo akuyona akuyona abalandeli abalandeli, kodwa futhi abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli abalandeli. Izici ze-insider-risk tracks. Ngokusekelwe ku-Proofpoint's , "Uma izigidi ezimbini ze-CISOs zihlanganisa ukuphazamiseka kwebhizinisi elidlulileyo ngonyaka elidlulileyo, kanye nezinkinga ze-insider-driven zihlanganisa isikhwama esifundeni." Voice of the CISO Report Umhlahlandlela we-CISO Report Ukusuka kwedatha kwenziwe ngokushesha. Ngaphandle kokusuka ngokushesha ngokusebenzisa ama-channels ezivamile ze-approval, idatha asebenza ngokushesha phakathi kwamakhemikhali, izixhobo, kanye namasipho we-AI. Ngokwenza lokhu, abacwaningi be-security zihlanganisa ukufinyelela kwezemikhali zabo. Izinqubo ze-Proofpoint zibonisa ukuthi e-U.S. zihlanganisa ukuthi idatha ye-customer lithunyelwe ngokusebenzisa i-publ generative AI platforms. 80% of CISOs I-80% ye-CISOs Umdlali ungayibiza umdlali we-AI imibuzo elula. Umdlali uxhumane idatha kusuka ku-system internal (isib. CRM, ERP) kanye emithonjeni amaningi e-external (isib. Iwebhu, i-social media), bese uxhumane imiphumela ehlanganisiwe kumalungu wonke umdlali we-collaboration platform ye-user (Slack). I-AI njenge-Data Movement Layer I-AI ikakhulukazi ukuguqulwa kwe-life cycle ye-data ngaphakathi kwezinhlelo zokusebenza. Ngokuvamile, kukhona i-handoffs evamile, lapho umdlali wabhala ifayela, ukuguqulwa, futhi wabhala umdlali. Amadivayisi we-Security angakwazi ukucubungula yonke iminyango ye-process. I-AI workflows ayinalo efanayo. Umdlali uthatha imibuzo. Umdlali we-AI uthatha ulwazi kusuka emithonjeni eziningi ukuze ufunde imibuzo. It ukunciphisa imibhalo, ukuguqulwa kwama-data, futhi ikhiqiza imiphumela enokuthunyelwe ku-collaboration platforms. Yonke iminyathelo ivela eminyakeni nge-supervision emininzi yabantu. "Kuba izinhlelo zokusebenzisa ama-agents ezisemthethweni abakwazi ukuhlola, ukhiye ikhodi, kanye nokusebenza phakathi kwezinhlelo eziningi, ukunambitheka kwenziwa umngciwane we-risk multiplier," wabhala abacwaningi we-Proofpoint. . related co-authored post Imininingwane Post Co-Author I-agent ye-AI ne-plugins ingasetshenziselwa kwama-system angaphezu kuka-one. Zihlanganisa idatha kusuka ku-CRM yakho, zihlanganisa ku-archives ye-imeyili, zihlanganisa ulwazi kusuka ku-collaboration platform, kanye nokuthumela imiphumela ku-analytics tool engaphandle kwe-company yakho. Kuyinto ezine kuya ku-data moves eminyakeni eyodwa yokusebenza okuzenzakalelayo. Ukukhishwa kwe-Insider: Intent vs. Convenience I-insider threat model eyadlulayo yayisekelwe ku-intention. Uyazi be-bad actors ngaphakathi kwebhizinisi idatha, futhi abasebenzi abalandeli abalandeli abalandeli ukwakhiwa kwebhizinisi. I-AI iye yandisa isisindo kusuka ku-intention kuya ku-comfort. I-Customer is not trying to cause harm. They’re trying to work more efficiently and meet tomorrow’s deadlines. Ukubeka i-Customer Contract ku-ChatGPT futhi ukhangela isibuyekezo kuyinto engcono kunokufunda ngesandla nge-40 amakhasi. Ukubuyekeza isibuyekezo se-financial ku-AI assistant ikhiqiza i-charts e-second instead of hours. Wonke isixazululo ibonise isivinini kunezinhlelo. Uninzi lwabasebenzi akuyona lapho izilinganiso zihlanganisa. Izinkampani zihlanganisa ngokushesha ukuthi abasebenzi we-data angakwazi ukunikela izixhobo ze-AI. Abanikeze noma yintoni okufanayo idatha uma ifakwe ku-AI uhlelo. Uma izinsizakalo zihlukile, abantu akufanele ukwenza lokho ukuze uthole umsebenzi yabo. acceptable AI use inthe workplace ukusetshenziswa kwe-AI ku-inthe workplace I-Collaboration Platforms njenge-AI Gateways I-Slack, i-Teams, i-Gmail, i-Google Workspace, ne-Microsoft 365 ziquka i-"first hop" ye-data njengezinto ezivela emkhakheni yakho yebhizinisi futhi isetshenziselwa izixhobo zokusebenza ze-AI. Izinzuzo ze-OAuth kanye nezixhumanisi ze-API zihlanganisa ukuxhumanisa. Umdlali ufakele i-plugin yokukhiqiza okuvumela ukuxhumana kwe-Slack noma ukunakekela ukuxhumana kwe-imeyili. I-app ihhovisi ukufinyelela okungaphezulu kwedatha ye-workspace. Iningi abantu abacindezela ku-"allow" ngaphandle kokufundisa lokho. Ngaphezu kwalokho, inkonzo ye-AI ye-third-party ingathola ama-messages, ama-file e-shared, ne-calendar information. I-CISOs inesibopho emikhulu ekwazi ukubona okuholela kulezi zokusebenza kwezi zokusebenza ze-workflows. Abaningi awukwazi ukujabulela imibuzo elula mayelana nezinsizakalo ze-AI ezisebenzisa abasebenzi ukuxhumana. Abanikeze isizukulwane esizayo sokuhamba ukuthi idatha esuka ngokusebenzisa lezi zokusebenza. Abanikeze ukuthi umuntu uxhumane ulwazi oluthile ku-AI izinhlelo noma lapho i-outputs zithunyelwe ngemuva kokufakelwa. Yintoni i-Controls ye-Traditional Amasipala Ngaphandle Izinkqubo zokhuseleko ze-legacy ziye ziye ziye ziye ziye ziye zihlanganisa imodeli ezingaphezu kwe-threat. Ngenxa yalokho, zihlanganisa izinzuzo ezithakazelisayo ekusebenzeni kanjani i-AI ikakhulukazi ukwandisa idatha phakathi kwezinhlangano ezintsha. 
 
 
 
 I-static vs. dynamic: I-DLP kanye ne-CASB ezivamile zihlanganisa ku-static governance, okuyinto isekelwe ngezinsizakalo ezivamile mayelana nokuthumela idatha. I-AI ibonise i-fluid kanye ne-multi-step workflows lapho idatha isithunyelwe kanye nokuthunyelwe ngaphakathi kwezinye imizuzu. I-speed kanye ne-variability ye-AI workflows zibonisa izinsizakalo ze-static futhi engabonakali. Umntu vs. okuzenzakalelayo: Ngezinyanga ezidlulile, ukwelashwa kwe-insider kulindiza abantu ukwenza izixazululo ngezinyathelo ezininzi. I-AI kusebenza ngokugcwele ukuchithwa, ukuguqulwa kanye nokuguqulwa kwedatha ngokusebenzisa ama-agents kanye / noma ama-assistants. IAM frameworks inikeza ukufinyelela kubasebenzisi kodwa ayenzelwe ukulawula ukuthi izimpendulo ze-AI okuzenzakalelayo zokusebenza kumakhasimende. I-Predictable vs. Emerging: Uma izindlu zokusebenza zihlanganisa izakhiwo zokuphathwa, amaqembu zokuphathwa zokusebenza ngokuvamile zihlanganisa izindlela zokuphathelela. Izindlu zokusebenza ze-AI zihlanganisa; umdlali angakwazi ukufinyelela idatha kusuka ku-5 izinhlelo ezihlukene ngezindlela engathengiswa. Kwakukwazi ukwenza izicelo zokuphathelela zonke izakhiwo ezitholakala zebhizinisi. Yini I-CISOs Imininingwane (Ou Ukulandelana Ngisho) Iziqu zeengcali ze-security zitholela ukunikezela lapho umkhakha uye. Lezi zihlanganisa kuphela ama-theoretical; zihlanganisa izivakashi ze-real-world ezihlangene nokulawulwa kwe-strategic. 
 
 
 
 
 
 Yini izixhobo ze-AI ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye ziye zibe. Izigaba zebhizinisi zingatholakala ku-AI workflows? Amakhasimende akufanele ukunikezele ukuthi izinhlelo zebhizinisi zebhizinisi zingathola ama-customer records, ama-financial data, i-source code, noma ama-data eyenziwe nezinsizakalo. Umthombo we-AI uhamba lapho? I-AI ivela ama-summary noma ama-rapports ezahlukile nabanye amaqembu noma ama-partners e-external. Kuyinto enhle kakhulu ukucubungula i-output njenge-input. Ungayifaka kanjani, ukucubungula kanye nokubhalisa ukuxhumana kwe-AI? Ngaphandle kwamandla yokubhalisa, amabhizinisi we-Security ayikwazi ukucubungula izikhangiso noma akuyona ama-records ezivela kumadokhumenti we-compliance lapho baqala ukuhlola imibuzo. Ukulungiselela ukuthi unemibuzo yokulawula umthetho (isib. I-CISO, i-Data Teams, i-legal, noma i-compliance) kusiza ukunciphisa izixazululo lapho umuntu akufanele. I-Industry Outlook ku-Insider Risk I-Insider risk akuyona kuphela inkinga elilodwa yabantu; kuya kuhlanganise ukuhlola izindlela ze-AI workflows kanye nezimo zokuhamba idatha. Ukuhlolwa kokusebenzisana kwamakhasimende akuyona kuphela akunakusiza amabhizinisi we-security ukuhlangabezana oku. Ama-agents e-automated, ama-API connectors, kanye ne-integrations nge-platform ye-collaboration. "Izinhlelo zikhala ukwelashwa ama-human signals, idatha ye-identity, kanye nama-events zobuchwepheshe njengama-streams ezahlukile," i-Proofpoint ibonise. "I-evolution elandelayo ye-insider risk management iyatholakala ku-connecting lezi zindawo, ngenxa yokuba i-risk efanelekayo ikhona emangalisayo e-dimension eyodwa." Ngenxa yalokhu, ukusetshenziswa kwe-AI uzodinga inkampani lonke ukuxhumana izibopho. Uma kuxhomekeke ukulawula ukulawula idatha, ukhuseleko lokusebenza, izicelo zomthetho, kanye nezidingo zokuxhumana, i-CISOs ayikwazi ukulawula le ngxaki ngokufanayo. Qaphela ukuthi izinkqubo zokuxhumana zihlanganisa ukuthi ama-Chief Data Officers ne-Security Leaders zokusebenza ngokufanayo ku-AI-specific governance, ngaphandle kokusebenza ezisodwa. Waze I-AI ayenza ama-insiders ezintsha. Kubalulekile ukucubungula indlela abantu abasebenzise idatha usihlanganisa ngosuku zonke. Ngaphandle kokufunda mayelana nokufundisa izinzuzo, isikhokelo lihlanganisa ukwenza lezi zindlela ngokuvumelana nezinsizakalo ezisebenzayo. 
 
 Lesi sihloko lithunyelwe njenge-release nguJon Stojan ngaphansi kwe-HackerNoon's Business Blogging Program. Lesi sihloko lithunyelwe njenge-release nguJon Stojan ngaphansi kwe-HackerNoon's Business Blogging Program.

The is an opinion piece based on the author’s POV and does not necessarily reflect the views of HackerNoon.

Lo msindo ukhiqizwa ngolimi lokuqala lwendaba!

The New Insider Risk: AI ukuguqulwa indlela idatha ihlukanise ngaphakathi inkampani

About Author

IMIBONO

HANG TAGS

LESI SIHLOKO SETHULWE NGAPHAKATHI

Related Stories

Zulu Network: Moving The Bitcoin Economy Forward With a Two-Tiered Bitcoin Layer 2 Architecture

The Billionaires F*cked Up

Zulu Network: Moving The Bitcoin Economy Forward With a Two-Tiered Bitcoin Layer 2 Architecture

The Billionaires F*cked Up

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps