On the second day of Pwn2Own Berlin 2026, researchers earned $385,750 by exploiting 15 zero-day vulnerabilities including successful attacks on Microsoft Exchange and Windows 11. The demonstrations underscore risks in enterprise and AI products while giving vendors 90 days to issue patches after public disclosure.

During day two of Pwn2Own Berlin 2026, participants secured $385,750 in awards by demonstrating 15 distinct zero-day vulnerabilities across various products such as Windows 11, Microsoft Exchange, and Red Hat Enterprise Linux for Workstations.
The Pwn2Own Berlin 2026 contest runs at the OffensiveCon conference between May 14 and May 16, concentrating on enterprise technologies together with artificial intelligence. Researchers stand to gain more than 1000000 dollars in total cash and rewards through successful attacks on up-to-date, fully patched systems in areas covering web browsers, enterprise applications, cloud-native and container setups, virtualization platforms, local privilege escalation, servers, local inference, and LLM categories.
Event regulations specify that every device under test uses the newest operating system releases, with all submissions required to achieve full system takeover and show arbitrary code execution. After disclosure at the competition, manufacturers receive 90 days to develop and release fixes for the reported flaws.
A standout performance came from Cheng-Da Tsai, known as Orange Tsai, with the DEVCORE Research Team, who received $200,000 for linking three separate issues that delivered remote code execution at SYSTEM level against Microsoft Exchange. Siyeon Wi obtained $7,500 by leveraging an integer overflow to breach Windows 11, while Ben Koo from Team DDOS gained root access on Red Hat Enterprise Linux for Workstations and took home a $10,000 prize. Separately, 0xDACA and Noam Trobishi took advantage of a use-after-free vulnerability in the NVIDIA Container Toolkit.
Within the AI track, Le Duc Anh Vu from Viettel Cyber Security earned $30,000 for compromising the Cursor AI coding assistant, Sina Kheirkhah representing Summoning Team showcased an OpenAI Codex flaw worth $20,000, and Compass Security demonstrated an exploit against Cursor for $15,000.
On the opening day, Orange Tsai had already claimed an additional $175,000 by combining 4 logic flaws that enabled a sandbox escape in Microsoft Edge. Valentina Palmiotti, operating under the name chompie for IBM X-Force Offensive Research, picked up $20,000 for achieving root on Red Hat Linux for Workstations plus $50,000 for a zero-day in the NVIDIA Container Toolkit.
Windows 11 also fell three separate times during that initial day to Angelboy and TwinkleStar03 collaborating with the DEVCORE Internship Program, Kentaro Kawane from GMO Cybersecurity, and Marcin Wizowski; each researcher walked away with $30,000 for their fresh local privilege-escalation zero-days.
Day three of the competition will see attempts against Microsoft Windows 11, VMware ESXi, Red Hat Enterprise Linux, Microsoft SharePoint, along with multiple AI coding tools.
During the previous Pwn2Own Berlin event, Trend Micro’s Zero Day Initiative disbursed 1078750 dollars across 29 zero-day reports and several collisions.
Enterprise teams running Exchange and Windows 11 should track these disclosures and prepare for patches within the stated 90-day vendor timeline to limit exposure in production environments.
Tap a lens to see what this story means for you.
Liked this? The Brief brings you the whole day in tech, verified, every morning.
Two minutes, free forever. What's in The Brief →
See what’s happening right now
The Feed runs all day — short, verified briefs the moment they break.
Open the FeedFollow @thecircuitry_
Every story we publish, as it happens. No noise between.
Reader-supported
The Circuitry is a passion project I've always wanted to build, and I love the work behind it.
Running it costs real money. APIs, hosting, time. To keep improving the site and growing this into something useful for everyone, those costs have to be covered.
Any contribution is appreciated. If not, no pressure. Thanks for reading.
OpenAI has cut the average success rate of attacks that extract protected reasoning from an earlier request to 0.2% on GPT-6.2 Astra, from 54.0% on GPT-6 Astra. Current-request attacks are at 0.9%, and selected safeguard improvements are being applied back to GPT-6 Astra.
A suspected China-linked group exploited VMware vCenter bug CVE-2026-123890 in early September, before the flaw was disclosed on Nov. 19. Federal agencies have until Nov. 23 to apply the vendor's mitigations for the known-exploited flaw.
Autonomous Cairn agents found an unpatched flaw in a medical imaging app and took control of a server in a university campaign that began September 8, 2026. Amazon rates a Chinese-nexus link as moderate confidence, and the project advisory still lists no patched version.
Google has linked a late-2025 data theft at a US organization to a 2026 intrusion at a second US organization by the Russian group COLD RELIC. Operators used stolen staff records from the first breach to impersonate an employee and phish the next target.
Microsoft has opened Ignite 2026 with Copilot Workspaces for people and agents, Microsoft IQ, and Cobalt 200 virtual machines that are generally available today. The new machines run up to 50% faster than the prior generation, and a Salesforce plugin in Copilot preview includes more than 38 sales skills.