Pemula Israel mengumpulkan sekurang-kurangnya $577.5 juta pada bulan Ogos kerana keselamatan AI mengetuai pembiayaan

Apa yang berlaku

CTech reports that Israeli startups raised at least $577.5 million across 12 officially announced funding rounds in August, plus an undisclosed strategic investment in cybersecurity startup Above. The outlet says August 2026 was the strongest August for Israeli startup funding since 2022, despite the month traditionally being quieter because of summer holidays. AI security was the dominant theme, with companies targeting risks created by autonomous agents and AI-powered attacks.

CTech reports that Israeli startups raised at least $577.5 million across 12 funding rounds officially announced during August, with a 13th investment announcement involving cybersecurity startup Above that did not disclose an amount. The outlet describes the result as especially notable because August is traditionally one of the quietest months for Israeli technology funding, when much of the industry slows for summer holidays. According to CTech, August 2026 was the strongest August for Israeli startups since 2022. The article is a roundup of funding announcements rather than an independently audited market-wide investment report, so the total depends on CTech’s stated inclusion criteria and the amounts publicly announced by the companies involved.

CTech reports that the largest round was Alice’s $140 million raise at a valuation of between $700 million and $800 million. The company is applying intelligence on malicious online behavior to security for AI models and agents, according to the outlet. Zenity raised $125 million in a Series C, bringing its reported total funding to $185 million, as it expands its work securing autonomous AI agents. Oligo Security and Corma each raised $60 million. CTech says Oligo is focused on runtime security amid faster AI-assisted vulnerability discovery, while Corma is developing a foundation model for defensive cybersecurity. The source does not independently confirm the companies’ technical performance, customer numbers or the valuation basis for Alice.

CTech also lists several smaller or adjacent rounds. Attestable raised $20 million to develop zero-knowledge cryptography for verifying activity inside AI systems, while June AI raised $20 million in pre-seed funding to build agents for software implementation work. Arrakis Security raised $8 million to monitor what autonomous AI systems do after gaining access to enterprise systems. Xpander raised $7.5 million for infrastructure intended to help companies deploy AI agents across models, clouds and workflows. Above received an investment from CrowdStrike, but CTech reports no dollar amount. The article gives no independent assessment of whether these products are in broad commercial use or how they compare with established security controls.

Beyond security, CTech reports that travel fintech Faye raised $50 million in a Series C at an estimated $500 million valuation and plans to use AI to automate insurance claims while expanding globally. QuantHealth raised $45 million in a Series B for AI simulations of clinical trials; the company says its technology has simulated more than 600 trials and is intended to help drugmakers test decisions before enrolling patients or committing large development budgets. Remepy raised $36 million in a Series A as it prepares its AI-driven Hybridopa Parkinson’s treatment for Phase III and expands into rare tumors through a partnership with Merck. CTech reports the company’s claim that the program improved motor and non-motor symptoms in Phase IIa trials, but the source does not provide trial data or independently verify those results.

Butiran sumber: calcalistech.com ↗

Mengapa ia penting

The funding pattern suggests investors are placing significant value on security, verification and deployment infrastructure as companies move AI systems from experiments into production. CTech reports major rounds for Alice, Zenity, Oligo Security and Corma, alongside smaller investments in Attestable, June AI, Arrakis Security and Xpander. The source does not independently verify the startups’ product performance, customer adoption or projected market size.

The funding concentration matters because it reflects a practical problem created by the expansion of AI: systems that can act with less direct human intervention introduce new questions about access, behavior, monitoring and accountability. CTech reports that investors backed companies securing AI models and agents, defending against AI-powered cyberattacks, monitoring production environments and helping enterprises move from pilots to deployment. These are operational concerns rather than speculative applications. If the startups’ approaches work, they could give organizations additional ways to control systems that can make decisions or take actions after receiving access.

CTech’s account highlights a shift in the security boundary. Traditional security programs often emphasize identifying vulnerabilities before software is deployed or controlling which users can access systems. The companies in the roundup are pursuing other points of control: Oligo focuses on runtime protection, Arrakis on autonomous agent behavior after access, Zenity on agent security, and Attestable on cryptographic verification. Corma is pursuing a defensive cybersecurity model. The source presents these as company strategies and investor bets, not as established evidence that a separate AI-security layer is always necessary or that these products prevent successful attacks.

Jumlah tersebut juga menunjukkan bahawa infrastruktur AI dan pengurusan risiko sedang dibiayai bersama aplikasi AI yang lebih jelas. CTech melaporkan pelaburan dalam infrastruktur penggunaan model silang dan merentas awan Xpander dan ejen pelaksanaan perisian AI Jun, merangka pelaksanaan perusahaan sebagai halangan kepada penggunaan yang lebih luas. Gabungan itu mencadangkan pelabur melihat kebolehpercayaan penggunaan, pengawasan dan penyepaduan sebagai peluang komersial dalam hak mereka sendiri. Walau bagaimanapun, artikel itu tidak melaporkan kiraan pelanggan, hasil, pengekalan, penilaian bebas atau insiden yang didokumenkan yang diselesaikan oleh mana-mana syarikat yang dibiayai. Oleh itu, ia menyokong kesimpulan tentang keutamaan pelaburan, bukan tentang kepimpinan pasaran atau keberkesanan produk.

Kepentingan awam yang lebih luas masih tidak pasti. Lebih banyak pembiayaan untuk alat pertahanan boleh meningkatkan keupayaan syarikat dan kerajaan untuk memantau sistem AI, mengesahkan tindakan dan bertindak balas terhadap serangan. Ia juga boleh mewujudkan pasaran yang sesak di mana produk yang bertindih menjadikan program keselamatan lebih sukar untuk dinilai. CTech tidak melaporkan sama ada teknologi syarikat pemula telah diuji terhadap persekitaran perusahaan biasa, diaudit secara bebas, diluluskan oleh pengawal selia atau terbukti mengurangkan bahaya. Jurang tersebut penting kerana tuntutan keselamatan sukar untuk dibandingkan tanpa ujian biasa dan hasil dunia sebenar yang didedahkan.

Apa yang perlu ditonton seterusnya

The most important unknowns are whether these companies can convert funding into widely deployed products, whether AI-specific security tools address risks that existing cybersecurity systems cannot, and whether specialized defensive models outperform general-purpose systems in real-world incidents. CTech also reports substantial funding for AI-enabled travel insurance and clinical-trial simulation, but the source does not establish regulatory approval, clinical outcomes or commercial results from those efforts.

Isu pertama yang perlu diperhatikan ialah sama ada syarikat yang dibiayai beralih daripada konsep keselamatan yang menjanjikan kepada penggunaan yang boleh diukur. CTech melaporkan bahawa Zenity sedang mempercepatkan pengembangan global, Faye merancang pertumbuhan global dan Xpander sedang membina untuk penggunaan merentas model, awan dan aliran kerja. Sumber itu tidak menyatakan bilangan pelanggan yang menggunakan produk ini, sejauh mana ia tersedia atau bila syarikat menjangkakan sistem mereka mencapai penggunaan pengeluaran yang lebih luas. Bukti seperti penempatan pelanggan yang disahkan secara bebas, insiden yang didedahkan, penilaian luaran dan perubahan yang didokumenkan dalam kadar serangan atau tindak balas akan menjelaskan sama ada pembiayaan itu diterjemahkan kepada perlindungan praktikal.

Soalan kedua ialah sama ada kawalan khusus AI menyediakan keupayaan yang kekurangan alat keselamatan siber sedia ada. Pendekatan keselamatan masa jalan Oligo, tumpuan Arrakis pada tingkah laku autonomi dan penggunaan kriptografi pengetahuan sifar oleh Attestable menangani bahagian berlainan masalah yang diterangkan oleh CTech. Outlet itu juga melaporkan rancangan Corma untuk membina model asas keselamatan siber defensif. Penilaian yang bermakna memerlukan butiran tentang ancaman yang diuji, positif palsu, kelajuan tindak balas, kos penyepaduan dan prestasi terhadap serangan yang menggunakan AI. Tiada satu pun daripada perbandingan tersebut disediakan dalam sumber, jadi tuntutan tentang lapisan keselamatan baharu kekal berpandangan ke hadapan.

Pelaburan penjagaan kesihatan dan insurans memerlukan penelitian mereka sendiri. CTech melaporkan bahawa Faye berhasrat untuk mengautomasikan tuntutan insurans dan QuantHealth mengatakan ia telah mensimulasikan lebih daripada 600 ujian klinikal. Remepy sedang menyediakan program rawatan dipacu AI untuk Fasa III dan telah melaporkan peningkatan dalam Fasa IIa. Perkembangan ini boleh menjejaskan keputusan yang melibatkan tuntutan, pembangunan ubat dan rawatan pesakit, tetapi sumber itu tidak memberikan status kawal selia, keputusan percubaan klinikal, kadar ralat, keperluan semakan manusia atau bukti bahawa hasil simulasi meramalkan hasil dunia sebenar. Butiran tersebut akan menjadi penting sebelum menilai manfaat praktikal atau awam.

Akhir sekali, jumlah pembiayaan harus dijejaki terhadap pembiayaan kemudian, penutupan, pemerolehan dan keputusan komersial. Gambar Ogos CTech menangkap pelaburan yang didedahkan, bukan kesihatan jangka panjang sektor AI Israel atau kejayaan mana-mana syarikat individu. Ia juga tidak menentukan sama ada keselamatan AI akan kekal sebagai tema pelaburan utama, sama ada penilaian yang dilaporkan akan bertahan, atau sama ada syarikat pemula boleh bersaing di peringkat antarabangsa. Pelaporan masa depan harus mengesahkan jumlah yang didedahkan, mengenal pasti bukti bebas penggunaan dan prestasi, dan membezakan penempatan yang telah siap daripada rancangan syarikat.

Panduan & kuiz berkaitan