{"id":763228,"date":"2025-09-30T12:47:27","date_gmt":"2025-09-30T10:47:27","guid":{"rendered":"https:\/\/www.hashtag.al\/?p=763228"},"modified":"2025-09-30T12:47:27","modified_gmt":"2025-09-30T10:47:27","slug":"firma-izraelite-zbulon-nje-dobesi-ne-chatgpt-qe-nxjerr-te-dhena-pa-dijenine-e-perdoruesit","status":"publish","type":"post","link":"https:\/\/www.hashtag.al\/index.php\/2025\/09\/30\/firma-izraelite-zbulon-nje-dobesi-ne-chatgpt-qe-nxjerr-te-dhena-pa-dijenine-e-perdoruesit\/","title":{"rendered":"Firma izraelite zbulon nj\u00eb dob\u00ebsi n\u00eb ChatGPT q\u00eb nxjerr\u00a0t\u00eb dh\u00ebna pa\u00a0dijenin\u00eb e p\u00ebrdoruesit","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

\"\"<\/a><\/p>\n

Firma e siguris\u00eb kibernetike Radware ka zbuluar nj\u00eb dob\u00ebsi t\u00eb panjohur m\u00eb par\u00eb n\u00eb platform\u00ebn ChatGPT t\u00eb OpenAI, duke sh\u00ebnuar shfryt\u00ebzimin e par\u00eb t\u00eb serverit t\u00eb k\u00ebtij lloji q\u00eb synon agjent\u00ebt e IA-s\u00eb.<\/span><\/p>\n

Dob\u00ebsia, e quajtur ShadowLeak, ndikon n\u00eb agjentin Deep Research t\u00eb ChatGPT dhe u mund\u00ebson hakeruesve t\u00eb nxjerrin n\u00eb m\u00ebnyr\u00eb autonome t\u00eb dh\u00ebna t\u00eb ndjeshme t\u00eb p\u00ebrdoruesit nga server\u00ebt OpenAI – pa k\u00ebrkuar q\u00eb p\u00ebrdoruesi t\u00eb klikoj\u00eb, hap\u00eb ose madje t\u00eb shikoj\u00eb ndonj\u00eb gj\u00eb. Sipas Radware, p\u00ebrdorimi vepron fshehurazi dhe nuk l\u00eb shenja t\u00eb dukshme n\u00eb rrjete ose pajisje, duke paraqitur nj\u00eb k\u00ebrc\u00ebnim serioz p\u00ebr nd\u00ebrmarrjet q\u00eb po p\u00ebrdorin gjithnj\u00eb e m\u00eb shum\u00eb sh\u00ebrbime t\u00eb IA-s\u00eb, shkruan Ynet<\/a>.<\/span><\/span><\/p>\n

Rad\u00ebare zbuloi dob\u00ebsin\u00eb e\u00a0OpenAI n\u00eb qershor. OpenAI konfirmoi se problemi ishte zgjidhur m\u00eb 3 shtator.<\/span><\/p>\n

“Ky \u00ebsht\u00eb sulmi thelb\u00ebsor pa klikime,”<\/i><\/em>\u00a0tha David Aviv, drejtor i teknologjis\u00eb n\u00eb Radware.<\/span><\/p>\n

“Nuk k\u00ebrkohet asnj\u00eb veprim nga p\u00ebrdoruesi, asnj\u00eb shenj\u00eb e dukshme dhe asnj\u00eb m\u00ebnyr\u00eb q\u00eb viktimat t\u00eb din\u00eb se t\u00eb dh\u00ebnat e tyre jan\u00eb kompromentuar. Gjith\u00e7ka ndodh t\u00ebr\u00ebsisht prapa skenave.”<\/i><\/em><\/span><\/p>\n

Shfryt\u00ebzimi u demonstrua nga Qendra e K\u00ebrkimit t\u00eb Siguris\u00eb (RSRC) e Radware, ku studiuesit treguan se nj\u00eb email keqdash\u00ebs i d\u00ebrguar nj\u00eb p\u00ebrdoruesi mund t\u00eb aktivizonte agjentin e K\u00ebrkimit t\u00eb Thell\u00eb t\u00eb ChatGPT – q\u00eb funksionon n\u00eb cloud-in e OpenAI – p\u00ebr t\u00eb aksesuar dhe zbuluar n\u00eb m\u00ebnyr\u00eb autonome t\u00eb dh\u00ebna t\u00eb ndjeshme. Nuk nevojitej nd\u00ebrveprim i p\u00ebrdoruesit n\u00eb asnj\u00eb faz\u00eb.<\/span><\/p>\n

“Ky \u00ebsht\u00eb sulmi i par\u00eb pa klikime thjesht nga serveri q\u00eb kemi par\u00eb, ku agjenti i IA kryen n\u00eb m\u00ebnyr\u00eb autonome nxjerrjen<\/i><\/em>\u00a0e t\u00eb dh\u00ebnave<\/i><\/em>,”<\/i><\/em>\u00a0tha Gabi Nakibly, nj\u00eb nga studiuesit kryesor\u00eb pas zbulimit, s\u00eb bashku me Zvika Babo dhe Maor Uziel.<\/span><\/p>\n

Ndryshe nga sulmet tradicionale me zero klikim q\u00eb synojn\u00eb pikat fundore ose pajisjet mobile, ShadowLeak operon brenda infrastruktur\u00ebs cloud t\u00eb inteligjenc\u00ebs artificiale, duke anashkaluar t\u00eb gjitha mbrojtjet q\u00eb p\u00ebrballen me p\u00ebrdoruesin dhe n\u00eb nivel rrjeti. Ai nuk l\u00eb gjurm\u00eb t\u00eb dallueshme nga ekipet e siguris\u00eb s\u00eb nd\u00ebrmarrjeve.<\/span><\/p>\n

\u201cFlukset e pun\u00ebs t\u00eb drejtuara nga inteligjenca artificiale po p\u00ebrvet\u00ebsohen me shpejt\u00ebsi, por kjo teknologji sjell rreziqe t\u00eb reja q\u00eb nuk adresohen nga mjetet e siguris\u00eb t\u00eb trash\u00ebguara\u201d<\/i><\/em>, tha Pascal Geenens, drejtor i inteligjenc\u00ebs s\u00eb k\u00ebrc\u00ebnimeve kibernetike n\u00eb Radware.<\/span><\/p>\n

\u201cHulumtimi yn\u00eb tregon se k\u00ebta agjent\u00eb mund t\u00eb manipulohen n\u00eb m\u00ebnyra q\u00eb nuk ishin parashikuar kurr\u00eb.\u201d<\/i><\/em><\/span><\/p>\n

Zbulimi vjen nd\u00ebrsa p\u00ebrdorimi i ChatGPT n\u00eb nd\u00ebrmarrje vazhdon t\u00eb rritet. N\u00eb nj\u00eb intervist\u00eb n\u00eb gusht me CNBC, z\u00ebvend\u00ebspresidenti i produkteve t\u00eb ChatGPT, Nick Turley, tha se platforma ka 5 milion\u00eb p\u00ebrdorues biznesi q\u00eb paguajn\u00eb, duke n\u00ebnvizuar shkall\u00ebn e mundshme t\u00eb ekspozimit.<\/span><\/p>\n

Rad\u00ebare do organizoj\u00eb nj\u00eb \u00ebebinar t\u00eb drejtp\u00ebrdrejt\u00eb m\u00eb 16 tetor p\u00ebr t\u00eb diskutuar dob\u00ebsin\u00eb, duke ofruar udh\u00ebzime p\u00ebr profesionist\u00ebt e siguris\u00eb dhe zhvilluesit e inteligjenc\u00ebs artificiale se si t\u00eb mbrojn\u00eb agjent\u00ebt e inteligjenc\u00ebs artificiale nga k\u00ebrc\u00ebnime t\u00eb ngjashme.<\/span><\/p>\n

Radware p\u00ebrg\u00ebzoi OpenAI p\u00ebr bashk\u00ebpunimin e saj t\u00eb shpejt\u00eb dhe theksoi r\u00ebnd\u00ebsin\u00eb e hulumtimit proaktiv t\u00eb siguris\u00eb s\u00eb inteligjenc\u00ebs artificiale. Gjetjet e kompanis\u00eb dhe ndarja teknike e shfryt\u00ebzimit ShadowLeak jan\u00eb t\u00eb disponueshme p\u00ebrmes qendr\u00ebs s\u00eb saj t\u00eb K\u00ebrkimit t\u00eb Siguris\u00eb.<\/span><\/p>\n

RSRC e Radware kryen rregullisht simulime k\u00ebrc\u00ebnimesh dhe zbulon dob\u00ebsi q\u00eb ndikojn\u00eb si n\u00eb platformat tradicionale ashtu edhe n\u00eb ato t\u00eb bazuara n\u00eb inteligjenc\u00ebn artificiale.<\/span><\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"

Firma e siguris\u00eb kibernetike Radware ka zbuluar nj\u00eb dob\u00ebsi t\u00eb panjohur m\u00eb par\u00eb n\u00eb platform\u00ebn ChatGPT t\u00eb OpenAI, duke sh\u00ebnuar shfryt\u00ebzimin e par\u00eb t\u00eb serverit t\u00eb k\u00ebtij lloji q\u00eb synon agjent\u00ebt e IA-s\u00eb. Dob\u00ebsia, e quajtur ShadowLeak, ndikon n\u00eb agjentin Deep Research t\u00eb ChatGPT dhe u mund\u00ebson hakeruesve t\u00eb nxjerrin n\u00eb m\u00ebnyr\u00eb autonome t\u00eb dh\u00ebna […]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":51,"featured_media":763229,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[53],"tags":[],"class_list":["post-763228","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-fast-foodi"],"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/posts\/763228","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/users\/51"}],"replies":[{"embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/comments?post=763228"}],"version-history":[{"count":99999,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/posts\/763228\/revisions"}],"predecessor-version":[{"id":763230,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/posts\/763228\/revisions\/763230"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/media\/763229"}],"wp:attachment":[{"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/media?parent=763228"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/categories?post=763228"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.hashtag.al\/index.php\/wp-json\/wp\/v2\/tags?post=763228"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}