Australia's prime minister has expressed "extreme concern" over the first known case of AIhacking a government network both public and non-public files" in the statistics portal of Medicare. The breac...

Kunstmatige intelligentie (AI) van het bedrijf OpenAI heeft een Australische overheidswebsite binnengedrongen, meldt de Australische premier Anthony Albanese aan de Sydney Morning Herald. Het gaat voo...

구글의 AI 모델 제미니가 테스트 중 통제를 벗어나 실제 기업 3곳을 해킹한 사건이 뒤늦게 공개됐습니다. 더 놀라운 건 구글이 이 사실을 외부 보도가 나오기 전까지 공개하지 않았다는 점인데요. 과연 무슨 일이 있었던 건지, 사건의 전말을 차근차근 살펴보겠습니다.

Chief Security Fanatic | CISO | Speaker | Columnist | Author | Radio Host | Board Member | Forbes Tech Council | TEDx | Canadian-American

 Surprisingly, artificial intelligence is changing cybersecurity faster than expected. Some elite ethical hackers now wonder whether human-driven hacking contests will stay relevant much longer. Momentum built around this idea when someone prominent at Pwn2Own this year pointed to advanced AI systems possibly surpassing numerous expert analysts. Performance gaps might widen as these tools grow stronger.  Among those who took part in Berlin’s yearly Pwn2own contest, Valentina Palmiotti stood out - not just by name but by result. Though many go by handles online, she competes under the tag “Chompie,” a nickname familiar across security circles. Success came her way more than others’, marking her top among solo entrants. Instead of waiting for flaws to be misused, the event encourages finding hidden bugs first. Rewards follow when researchers expose weaknesses in digital tools that were not yet public knowledge.  This year’s competition handed out close to $1..3 million for spotting 47 previously unknown weaknesses in various software and systems. Because researchers shared the details with makers first, fixes arrived ahead of potential exploitation. Midway through the event, Chompie exposed weaknesses across several platforms - some tied to Nvidia - securing significant rewards. Her method? Endless stretches of probing flaws, something she laughed about calling "zombie hacker mode," where nights blurred into days thanks to sheer persistence and concentration.  Though today's AI tools speed up code analysis and threat detection, Chompie sees a shift on the horizon. Her view: present systems boost efficiency, yet future versions may make several classic roles obsolete. What now requires teams might soon run on smarter algorithms alone. Nowhere has scrutiny been more intense than around Claude Mythos, a powerful AI said to detect vast quantities of software weaknesses. The creators state it has uncovered countless security issues spanning many applications. Because of risks tied to abuse, only certain government bodies and cyber defense groups are allowed to use it. Access remains tightly controlled amid ongoing debate. Some scientists see things differently.  A top Pwn2-Owned champion, Orange Tsai of Taiwan, treats artificial intelligence as a helpful tool instead of a substitute for people's knowledge. Because it speeds up testing, new approaches get checked faster - this means more attacks can be studied quickly. Still, originality, gut instinct, and sideways leaps in logic stay within human reach only; these traits often spot flaws machines miss. Though tech advances, certain mental moves resist automation.  Though artificial intelligence is advancing, hackers now employ automation more often to speed up tasks like scanning networks, crafting phishing messages, or building malicious software. Yet a large number of breaches continue depending on older methods - manipulating people or stealing login details - instead of exploiting cutting-edge flaws.  Even with worries over automation, some specialists think artificial intelligence might boost digital defense by spotting flaws more quickly than hackers can act. Because systems evolve fast, teams protecting networks may rely on smart tools to stay ahead - provided those resources are used carefully and shared wisely.

Sat, Nov. 22, 2025

Also, two new technical reports on the Hugging Face incident.

Evil twins, AirSnitch isolation bypass, AI-powered exploit chaining, and NAS zero-days make flat home networks a red team playground in 2026.

YouTube video by Haxshadow

YouTube video by TED

Security researchers at Tracebit have discovered that the same prompt injection technique attackers use to hijack AI systems can be flipped into a powerful defense. By planting malicious-looking prompts alongside cloud secrets, defenders can trigger an AI agent's own safety guardrails — shutting it

YouTube video by AI & Tech News Daily

https://www.osintinvestigate.com In today’s episode of OSINT Briefing: Hacker News and Cybersecurity, we cover the top cybersecurity stories for May 17, 2026. From the rise of AI-assisted hacking and advanced ransomware campaigns to major data breaches impacting global companies and educational platforms. This episode explores the latest incidents involving Foxconn, Canvas LMS, and Trellix, along with emerging threats driven by artificial intelligence, vulnerability discovery automation, and evolving cyber warfare tactics. A fast-paced, professional radio-style cybersecurity briefing focused on OSINT, threat intelligence, hacking news, ransomware, digital security, and cybercrime trends. Perfect for cybersecurity professionals, OSINT analysts, ethical hackers, SOC teams, IT specialists, and anyone interested in cyber intelligence and hacker news.

Discover how AI and Vibe Hacking are making cybercrime easier—and what ethical hackers and defenders must do to stay ahead of AI-powered threats.