Omer Ginosar showed a packed room data pros at #hayaData 2026 how it works: data products shipped as backend services on the platform R&D already runs, updates delivered in Slack, and an LLM that decides only whether something deserves someone's attention today.

𝐌𝐨𝐬𝐭 𝐚𝐝𝐯𝐚𝐧𝐜𝐞𝐝 𝐀𝐈 𝐦𝐨𝐝𝐞𝐥 + 𝐫𝐞𝐝 𝐭𝐞𝐚𝐦 = 𝐦𝐨𝐫𝐞 𝐬𝐞𝐜𝐮𝐫𝐞 ? Not necessarily. Turns out that a better model helps. But without the right context, it’s still guessing. #AISecurity #CloudSecurity #RuntimeSecurity

🎊 Virtual Patching is live. 🐝 Now you can protect a running applications from a specific vulnerability in one click - while engineering takes the time it needs to fix it. Read → hubs.li/Q04BMcMF0 #RuntimeSecurity #VirtualPatching #AISecurity #CloudSecurity

🚀 We’re now part of @Anthropic's Cyber Verification Program, with access to Claude Mythos for attack validation. If attackers are going to have that capability, defenders need it too. 💪 #AISecurity #RuntimeSecurity #CyberSecurity #Anthropic

Attackers have AI. Defenders have home advantage. 🐝 Our CEO Dror Kashti writes in @Calcalistech about turning deep runtime context into an advantage: proving real attack paths, fixing exposure and stopping harmful behavior as it happens. Read the piece: hubs.li/Q04BpyYc0

Cloud security’s math is changing. 🍬 Sweet’s Chief AI Officer, Yigael Berger, is taking the stage at Toronto’s AI x Cloud Security Summit to unpack how AI changes the cost and speed of finding, exploiting and fixing vulnerabilities. At #SECTOR26? Come say hi!

We're excited to share that Sweet Security has been recognized by Gartner® as a Sample Vendor for Application Detection and Response in the 2026 Gartner® Hype Cycle™ for Application Security! #ApplicationSecurity #ADR #CloudSecurity #AISecurity #RuntimeSecurity #CyberSecurity

Black Hat's a grind. This is the good kind. ☕ Watch reality change in real time, get caught in the act, and walk out with the shirt in one hand and a coffee in the other. 👉 Come see what we’ve been brewing hubs.li/Q04qb28Y0 #BlackHat #BHUS

We're bringing The Sweet Stop to Black Hat ☕ Aug 5–6 🍭 Mandalay Bay 🍭 Booth 5721 Barista-made coffee, live runtime demos & a first look at how Sweet stops bad behavior in runtime. 👉 Coffee's on us - reserve your 20-min session: hubs.li/Q04p2qcM0

🇺🇸 Happy 4th of July - 250 years of American independence! To our US team, customers, partners, and the security community: thank you for being part of our journey. Wishing you a safe and joyful holiday from all of us at Sweet! 🎆 #America250 #FourthOfJuly

🎙️ Our CEO & co-founder Dror Kashti joined @avivfrenkel podcast @calcalist. It's in Hebrew, but the themes translate 👇 30+ years in Israel's cyber establishment, why "agentic" isn't a dirty word & why he's happy to spend on AI tokens. 🎧 hubs.li/Q04mYWGk0

The gap between granted access and used access is where most identity risk hides. Introducing Sweet Entitlements: runtime-powered visibility into permissions, identities, and access paths. Read more: hubs.li/Q04lHM-y0

Big milestone for Sweet Security: we're expanding into the U.S. federal market and pursuing FedRAMP Moderate authorization, with @Coalfire supporting our journey. Learn more: hubs.li/Q04kSG320 #FedRAMP #CloudSecurity

Another read on Sweet Attack. 📰 TechDay US gets to the core of why we built it: fewer theoretical findings, more proof of which attack paths work. Sweet Attack exposes attack chains with evidence and remediation paths. Read more: hubs.li/Q04gGjYC0 #AISecurity

Sweet Attack is already making headlines. 🗞️ SecurityWeek covers the shift behind our launch: AI red teaming only works when the agent understands the real environment. Sweet Attack validates the paths attackers would actually take. Read the coverage: hubs.li/Q04gpp340

Attackers rarely move one finding at a time. They chain vulnerabilities, identities, APIs, permissions, and runtime behavior into usable paths. Sweet Attack validates exploitable attack paths, so teams know what to fix first. See how: hubs.li/Q04gshnS0 #RedTeam #Mythos

Mythos meets its match. Introducing Sweet Attack 🤖: the AI red-team agent powered by Sweet’s runtime intelligence to validate real attack paths in production. The myth is that defenders can’t keep up. Sweet Attack debunks it. Read now: hubs.li/Q04gpSTk0 #RedTeam #Mythos

Mythos meets its match. Tomorrow morning, 9am ET. This Wednesday.

Tomorrow. 6am ET. The answer to Mythos. Stay tuned.

Twelve consecutive years of clean red-team reports. Three days into our engagement: fully exploitable attack chains. On Wednesday — May 13 — we'll tell you how. Announcing soon.