The safety check that is supposed to stop an AI coding agent from running a dangerous command can be walked straight past using a shell trick that has been public for decades. New research from Adversa AI, which is named the bypass GuardFall, found it works against ten of the eleven popular open-source coding and computer-use agents the firm tested. Only one, "Continue," was built to defend against it. Why does it matter? These agents run shell commands with your full account access. Point one at a booby-trapped repository or software package, and a hidden instruction can quietly run a command that wipes files or steals the secrets your account can reach, from SSH keys and cloud credentials to anything sitting in your home folder. How does it get past the guard? Most of these agents try to stay safe by checking each command against a blocklist of dangerous patterns before running it. The flaw is that they check the command as plain text, while bash rewrites that text before it actually runs. The shell strips quotes and expands shortcuts, so the filter and the shell end up looking at two different things. The simplest example: a filter watching for rm sees nothing wrong with r''m, because to a text matcher those are different strings. Bash removes the empty quotes and runs rm anyway. The same idea works in other forms: a command hidden in base64 and piped into a shell, or ordinary tools like find and dd turned destructive with the right flag. The researchers call this not a bug but "a dangerous convention and a class of problems," which is why adding more blocklist patterns fixes none of it. There is no single CVE to track or patch. Two things have to line up for an attack to
Jul 1, 2026 · via thehackernews.com
AI agents seen reshaping cybersecurity Experts urge for major paradigm shift to active immunity at forum in Beijing By Ma Si | China Daily | Updated: 2026-07-01 00:00 As artificial intelligence agents evolve from assistive tools into autonomous decisionmakers and executors, the traditional logic of cybersecurity defense is being fundamentally rewritten, experts and officials said. Li Bing, vice-chairman of the All-China Federation of Industry and Commerce, said the AI industry is at a critical juncture, transitioning from isolated technological breakthroughs to full-scale value creation. He called on private enterprises — as key drivers of technological innovation — to actively participate in the "AI plus" initiative, increase research and development investment, and accelerate the large-scale deployment and commercialization of foundation models and agent technologies, while adhering to security guardrails and ensuring that AI benefits humanity. Li made the comments at the 14th Internet Security Conference, also known as ISC. AI 2026, held in Beijing last week. Ren Xianliang, secretary-general of the World Internet Conference, said: "The development of AI agents must embed values of fairness, inclusiveness and security throughout the R&D and application process, while strengthening international rule-making and multilateral coordination to make AI agents a global public good that safeguards security and benefits all." Zhao Zhiguo, vice-chairman of the Internet Society of China and former chief engineer of the Ministry of Industry and Information Technology, called for a holistic approach to national security, urging stronger core technology R&D, improved tiered classification management and assessment mechanisms, and the establishment of security standards covering the entire lifecycle from development and deployment, to operations and maintenance. Wu Hequan, an academician from the Chinese Academy of Engineering, said that AI agent-driven attacks now exhibit autonomous decision-making, swarm coordination and continuous evolution. He argued that the security paradigm must shift from passive defense to active
Jul 1, 2026 · via global.chinadaily.com.cn
Intrusion Inc (NASDAQ:INTZ) shares are up almost 43% during Tuesday’s premarket session. On Monday, the company announced its acquisition of VigilAigent to enhance its cybersecurity offerings significantly. The company expects the acquisition to add approximately $3.5 million in annual recurring revenue from multi-year contracts. The deal will also add over 80 reseller partners and an installed base of around 1,000 customers to expand the company’s commercial footprint. The acquisition of VigilAigent will integrate its proprietary Agentic AI engine, The Oracle, with Intrusion’s TraceCop database, creating a more robust cybersecurity platform. This strategic move will bolster Intrusion’s capabilities in threat detection and response, positioning the company for future growth. As of March 31, 2026, Intrusion’s cash and cash equivalents stood at $1.4 million. What Intrusion Inc Does: Cybersecurity And Threat Intelligence The recent acquisition of VigilAigent is significant as it not only adds revenue but also enhances Intrusion’s technological capabilities in the rapidly evolving cybersecurity landscape. With the integration of VigilAigent’s AI-driven solutions, Intrusion is well-positioned to expand its market reach and improve its service offerings. INTZ Earnings Date, Estimates And Analyst Price Targets Intrusion is slated to provide its next financial update on August 11, 2026 (estimated). - EPS Estimate: Loss of 10 cents (Up from Loss of 10 cents) - Revenue Estimate: $1.57 million (Down from $1.90 million) Analyst Consensus & Recent Actions: The stock carries a Buy rating with an average price target of $5.25. Recent analyst moves include: - Ascendiant Capital: Buy (Lowers Target to $9.50) (June 8) - Ascendiant Capital: Buy (Lowers Target to $10.50) (April 9) - HC Wainwright & Co.: Neutral (Lowers Target to $1.00) (March 30) INTZ Price Action: Intrusion shares were up 26.86% at $0.86 during premarket trading on Tuesday, according to Benzinga Pro data. Photo via Shutterstock © 2026 Benzinga.com. Benzinga
Jul 1, 2026 · via benzinga.com
New Microsoft research shows how attackers can hijack AI agents that act on a user's behalf, using nothing more than a poisoned tool description to make the agent quietly hand over company data to an outsider. The trick is that the agent never breaks a rule. Every step looks routine, so in a default setup no alarm may fire. The work comes from Microsoft Incident Response and its Defender security research team, and it lands as companies start letting AI do more than read and summarize. What changes when an agent can act Until recently, the workplace AI risk was mostly framed around what a model read and wrote. A poisoned document could skew an answer, and that was mostly where it ended. Agents are different. Microsoft 365 Copilot can send email, create files, and change calendars. Custom agents built in Copilot Studio or Azure AI Foundry can reach into business systems and run multi-step jobs on their own. The same injection trick that biases a summary now triggers an action. Against a reader, an attack changes the output. Against an agent, it changes what the software actually does. These agents reach business systems through MCP, the Model Context Protocol, an open protocol that lets an AI call outside tools the way an app calls an API. Microsoft calls it the fastest-growing part of the agentic AI supply chain, which makes it an expanding attack surface. How the attack works Every MCP tool ships with a description: a few lines of plain text that tell the agent what the tool does and when to use it. The agent reads that text to decide how to act. That is the whole weakness. The description is just words, and words can carry instructions. Microsoft walks through it with an invoice example, built
Jun 30, 2026 · via thehackernews.com
Collaboration enhances protection of patient and employee information PLANO, Texas, June 30, 2026 /PRNewswire/ -- Heart to Heart Hospice, one of the country's largest private providers of hospice care, has partnered with CyberMaxx to strengthen the organization's cybersecurity capabilities and further protect sensitive patient and employee information. CyberMaxx provides healthcare-focused cybersecurity expertise, continuous security monitoring, endpoint detection and response capabilities, investigation support, and defined escalation pathways. The partnership complements Heart to Heart's internal information technology and security teams while adding specialized expertise and enhanced response capacity. Heart to Heart Hospice operates across more than 80 clinical and administrative locations in multiple states. The organization supports more than 2,800 employees and clinicians serving more than 4,500 patients. Many members of their care teams work in patients' homes and rely on secure, reliable access to technology throughout the day. "Our information technology and security teams recognize the importance of protecting patient, employee, and company information without interrupting patient care," said David Ewers, Senior Director of Information Technology for Heart to Heart Hospice. "With that responsibility in mind, we partnered with CyberMaxx to strengthen protection across the organization." The partnership is part of Heart to Heart's broader commitment to technology, modernization, operational resilience, and responsible growth. Over the past 18 months, the organization has grown by approximately 25 percent while continuing to expand its services into additional communities. About Heart to Heart Hospice Heart to Heart Hospice, founded in 2003, is one of the largest private providers of hospice care in the United States. Headquartered in the Dallas-Fort Worth Metroplex, Heart to Heart provides a broad range of hospice services to patients with life-limiting illnesses across 80 locations in multiple states. At Heart to Heart, we understand both the physical and emotional challenges for our patients and their loved ones and our compassionate
Jun 30, 2026 · via prnewswire.com
- GAO has identified 10 remaining priority recommendations for NTIA - Recommendations address spectrum management, cybersecurity and broadband programs - The 2026 FedCiv Summit on Oct. 29 will examine AI, cloud and more The Government Accountability Office has called on the National Telecommunications and Information Administration to implement its remaining priority recommendations, saying the actions could improve the agency’s operations across spectrum management, IT modernization, cybersecurity and broadband programs. As federal agencies continue advancing modernization efforts, government and industry leaders will gather at the Potomac Officers Club’s 2026 FedCiv Summit on Oct. 29 to discuss artificial intelligence deployment, cloud, data and compute infrastructure, workforce enablement, and cross-agency and enterprisewide programs. Sign up now and join experts as they discuss technologies and strategies shaping modernization initiatives across the federal civilian sector. In a report publicly released Monday, GAO said NTIA has implemented one of the 11 priority recommendations identified in July 2025, leaving 10 priority recommendations open as of May. What Are GAO’s Recommendations to Manage RF Spectrum & Cybersecurity Risks? GAO recommended that NTIA, in consultation with the Federal Communications Commission and other relevant federal agencies, monitor and update its collaboration guidance for radio frequency spectrum management. It also recommended that NTIA establish shared procedures for designing spectrum studies used during domestic and international spectrum management activities to strengthen interagency collaboration and better position the United States at international spectrum conferences. For spectrum infrastructure cybersecurity, GAO said NTIA relies on multiple custom IT systems that are outdated despite awarding modernization contracts in 2024. GAO recommended that the agency implement key planning and cybersecurity practices as it modernizes those systems to better support the success of its IT modernization efforts. What Is GAO’s Recommendation Regarding Federal Broadband Programs? GAO recommended that NTIA improve how it communicates the financial sustainability of broadband
Jun 30, 2026 · via executivegov.com
Beyond the perimeter: Modernizing protective intelligence for national special security events Commentary Read more
AP Photo/J. Scott Applewhite, File Wyden blocks nominee to lead NSA and Cyber Command Federal Newscast Read more
Jun 30, 2026 · via federalnewsnetwork.com
Two researchers have found six security flaws in AirDrop and Quick Share, the wireless features that beam files between nearby devices with no cables or shared network. An attacker within wireless range, with just a laptop and no prior connection, can crash the sharing service on a Mac or iPhone set to receive from anyone, with no tap or prompt. The same research found Quick Share flaws that bypass Samsung's session checks and trigger a potentially exploitable crash in Google's Windows app. The two features run inside an ecosystem of more than five billion active Apple and Android devices, though the tested bugs hit specific implementations and versions. The work, laid out in a new research paper by Arash Ale Ebrahim and Nils Ole Tippenhauer of the CISPA Helmholtz Center for Information Security, is the first to pull both stacks apart side by side, above the radio layer, where discovery becomes session handling, parsing, and trust decisions. The fixes have already started. Apple has patched one of the three AirDrop bugs and assigned it a CVE, though the advisory is not yet public; the other two are still in coordinated disclosure. Google paid a bounty for the Windows flaw and has landed a code fix, with its CVE still pending. Samsung's two bugs were handed to Google and remain under investigation. No public reports of these flaws being exploited have surfaced as of this writing. Three ways to knock out Apple's sharing All three AirDrop flaws end in the same crash: they take down sharingd, the background service on macOS and iOS that handles AirDrop. The catch is that this service also runs AirPlay, Handoff, Universal Clipboard, Continuity Camera, and NameDrop, so one crash takes the whole set down together. The simplest of the three needs only a single malformed
Jun 30, 2026 · via thehackernews.com
Convince an AI browser that it is playing a game, and it can hand over your login details. That is the finding behind BioShocking, a technique from security firm LayerX that tricked six AI browsers and assistants into copying a user's credentials and sending them to an attacker. The targets included OpenAI's ChatGPT Atlas, Perplexity's Comet, and Anthropic's Claude browser extension. An AI browser is one that can act for you, not just read pages. Switch it to agent mode, and it can click, type, and reach into the sites you are already signed into. That access is the whole point, and it is also the problem. The trick works because of how these agents read. The web page and your own instructions arrive as a single stream of text. That lets a malicious page slip in commands dressed up as ordinary content or game rules, and the agent cannot reliably tell the difference. Researchers call this indirect prompt injection. How the trick works The attack starts with a web page built as a puzzle. To fit its dystopian theme, the puzzle rewards wrong answers, like insisting that 2 + 2 = 5. Once the agent accepts that "wrong" is the winning move, it follows game logic instead of safety logic. The final step of the puzzle asks it to grab the user's credentials, and not one of the six agents flagged that as something it should refuse. The dangerous part is where the agent looks. In the test, a link was sent to the victim's work GitHub repository, where it pulled SSH login credentials and passed them to the attacker. LayerX used a harmless plaintext file, but the same trick could point the agent at other resources it can reach in that session: open tabs, signed-in accounts, and internal
Jun 30, 2026 · via thehackernews.com
Breaches are being buried. AI is reshaping the risk landscape. Leaders and front-line practitioners are reporting different realities.
To uncover what's really happening inside organizations, Bitdefender surveyed 1,200 IT and cybersecurity professionals about the changes, pressures, and priorities defining cybersecurity today.
The full report goes deeper - more data, more peer benchmarks, more context for the decisions your team is navigating right now.
Jun 30, 2026 · via bitdefender.com
Cybersecurity In-Depth: Feature articles on security strategy, latest trends, and people to know. Why Identity Security Is Your Cyber Career Entry Point As AI reshapes cybersecurity workflows, John Paul Cunningham, CISO at SIlverfort, says the technology is creating opportunities rather than eliminating jobs — and there are more ways than ever to break into the essential field. At a Glance - Despite AI automation of certain tasks like log analysis, cybersecurity still requires human oversight - Leadership foundations and physical security expertise from military service naturally complement cybersecurity roles - Professionals can leverage their existing strengths and backgrounds to find a path that fits in cybersecurity As organizations grapple with emerging threats and transformative, new technologies, John Paul Cunningham, chief information security officer (CISO) at SIlverfort, says demand for skilled security practitioners shows no signs of slowing — despite concerns that artificial intelligence might diminish career prospects for newcomers. In this latest Heard It From a CISO video series installment, Cunningham brings his perspective to these industry dynamics. With a background spanning military service, financial sector duty, and now his role in protecting Silverfort's corporate infrastructure, Cunningham shares with us the multifaceted nature of modern security leadership. The conversation around AI's impact on cybersecurity jobs has reached a fever pitch, with junior analysts questioning whether automation will render their skills obsolete. Yet industry veterans like Cunningham argue that this technological shift represents evolution rather than elimination. While AI excels at parsing massive log files and identifying patterns, the human element remains irreplaceable in areas requiring strategic thinking, stakeholder engagement, and adaptive problem-solving. Perhaps most striking is the field's accessibility through multiple entry points. Unlike many professions that demand a singular educational path, cybersecurity welcomes talent from technical backgrounds — audit and compliance roles, project management, law enforcement, and more. This
Jun 30, 2026 · via darkreading.com
UAE Launches National Accelerator to Build Homegrown Cybersecurity Startups The program aims to produce a new generation of national cybersecurity companies capable of developing innovative solutions that meet the needs of both local and international markets. News - UAE Launches National Accelerator to Build Homegrown Cybersecurity Startups - Employee Resistance to AI Becomes UAE Firms' Top Workforce Risk: Report - Apple’s AI Hardware Strategy Faces Pressure on Two Fronts - Saudi Arabia Launches AI Tourism Vision, Unveils TourismX Platform to Power Growth - Ford Rehires 350 Engineers After Finding AI Alone Wasn't Enough - $725 Billion Later, 91% Say Organizations Still Aren't Realizing AI's Full Value The UAE is reinforcing its national cybersecurity commitment with the launch of a national accelerator program dedicated to cybersecurity startups. The Khalifa Fund for Enterprise Development (KFED), in partnership with the UAE Government Cybersecurity Council, has launched a specialized program to produce a new generation of national cybersecurity companies capable of developing innovative solutions that meet the needs of both local and international markets. Unveiled during an MZN Hub Al Ain event, in the presence of Dr. Mohamed Al Kuwaiti, Head of Cyber Security for the UAE Government, and KFED CEO Mouza Obaid Al Nasri, the program has been developed in collaboration with CyberE71, the country’s flagship cybersecurity startup incubator. “Cybersecurity is no longer merely a technical sector; it has become a strategic driver of the digital economy and one of the fundamental pillars of the security and sustainability of societies,” said Dr. Al Kuwaiti. Building upon the existing memorandum of understanding (MoU) between KFED and the council, the initiative aims to attract, develop, and support cybersecurity startups through a comprehensive approach. The framework will provide startups with specialized mentorship, partnership-building opportunities, and access to experts, investors, and key industry stakeholders. Meanwhile, CyberE71 will
Jun 30, 2026 · via mitsloanme.com
WASHINGTON (June 30, 2026) – According to Five Eyes spy agency, advanced AI models are improving quickly enough to outperform cybersecurity systems within months. This has increased the likelihood of a hack on these systems. Security agencies like Britain, the United States, Australia, Canada, and New Zealand are encouraging businesses to prepare themselves as AI evolves.
This concern also overlaps with national security. Earlier this month, Anthropic suspended access to their Mythos 5 model to comply with a U.S. national security order, proving that AI is becoming a security risk.
For more analysis on these developments, please consider Scott J. White, Associate Professor and Director of the Cybersecurity Program and Cyber Academy at the College of Professional Studies. White is an expert in cybersecurity, cybercrime, counter-terrorism and infrastructure protection. He has worked for a variety of law enforcement agencies in the US, the UK and Canada, holds a Queen’s Commission, and was an officer with the Canadian Security Intelligence Service.
To schedule an interview, please contact Nadia Payne at nadia [dot] paynegwu [dot] edu (nadia[dot]payne[at]gwu[dot]edu) or GW Media Relations at gwmediagwu [dot] edu (gwmedia[at]gwu[dot]edu).
-GW-
Jun 30, 2026 · via mediarelations.gwu.edu
About
Press
Copyright
Contact us
Creators
Advertise
Developers
Terms
Privacy
Policy & Safety
How YouTube works
Test new features
NFL Sunday Ticket
© 2026 Google LLC
Jun 30, 2026 · via youtube.com
Cybersecurity jobs available right now: June 30, 2026 AI Offensive Security Engineer AGAPI | UAE | On-site – View job details As an AI Offensive Security Engineer, you will leverage AI and LLMs to accelerate offensive security research, exploit development, vulnerability discovery, and security automation. You will validate AI-generated findings through manual testing, conduct authorized security assessments, and produce high-quality technical reports and remediation guidance. AVP, Enterprise Authentication & Directory Services Synchrony | USA | Hybrid – View job details As an AVP, Enterprise Authentication & Directory Services, you will lead the modernization of enterprise identity by driving the transition from Active Directory to Microsoft Entra ID, architecting secure hybrid identity solutions, managing large-scale IAM transformation programs, and governing Entra ID security, lifecycle management, application integration, and advanced access controls. CISO Lumafield | USA | On-site – View job details As a CISO, you will define and execute the organization’s cybersecurity strategy, oversee secure cloud architecture and product security, embed security across the SDLC, lead compliance and risk management programs, strengthen incident response, support customer security engagements, and foster a security-first culture. Get weekly updates on new cybersecurity job openings. Subscribe here! Cloud Security Engineer Spotify | USA | On-site – View job details As a Cloud Security Engineer, you will design and implement cloud security best practices, integrate security across the software lifecycle, conduct threat modeling and risk assessments, enhance cloud security tooling, support incident response, and drive security improvements across engineering teams. Cloud Security Network SME vSecureLabs | USA | On-site – No longer accepting applications As a Cloud Security Network SME, you will design and implement secure multi-cloud solutions across Azure, AWS, and GCP, architect hybrid connectivity and cloud networking, implement Zero Trust and cloud security best practices, and automate infrastructure deployment using Infrastructure as Code. Cybersecurity
Jun 30, 2026 · via helpnetsecurity.com
Hottest cybersecurity open-source tools of the month: June 2026 Presented here is a curated selection of noteworthy open-source cybersecurity solutions that have drawn recognition for their ability to enhance security postures across diverse settings. OWASP Agent Memory Guard: Stop AI agents from being weaponized through their own memory AI agents keep memory across sessions. Conversation history, vector stores, scratchpads, and RAG indexes persist between runs, and anything written into that store becomes a privileged input the agent reads back later. An attacker who plants text in the wrong field can override an agent’s instructions, pull out user data, or steer future tool calls, and the effect survives across sessions because the memory does. Agent Memory Guard is an open-source runtime defense layer that sits between an agent and its memory store, screening every read and write through a pipeline of detectors and a YAML policy. Agent Threat Rules: Open detection rule format for AI agent security threats AI agents run inside coding assistants, MCP servers, and multi-agent frameworks, and the access that makes them useful also opens paths to prompt injection, tool poisoning, and credential theft. Public CVE feeds carry agent-execution flaws that reach production faster than the tooling built to catch them. Agent Threat Rules, or ATR, is an open detection format aimed at this category of attack. AgentGG: Open-source agentic SAST scanner Static analysis tools have spent years matching source code against known-bad patterns and handing engineers long lists of candidate issues to triage by hand. AgentGG approaches the same job with AI agents that read the code, follow imports, walk the call graph, and confirm a finding before they report it. The project is an open-source agentic SAST scanner released under the Apache 2.0 license. DockSec: Open-source AI-powered Docker security scanner DockSec is an OWASP Incubator Project
Jun 30, 2026 · via helpnetsecurity.com
About
Press
Copyright
Contact us
Creators
Advertise
Developers
Terms
Privacy
Policy & Safety
How YouTube works
Test new features
NFL Sunday Ticket
© 2026 Google LLC
Jun 30, 2026 · via youtube.com
Over the last week, there's been a lot of swirl from a former employee alleging an insider threat within Huntress: passionate posts, public thoughts, and even some conspiracy theory. As the CEO of a company I also co-founded, I've had a front row seat to a volatile situation mixing emotional connections, nuances of ethics, partial truths, US and UK employment law, and active investigations, all while maintaining empathy for all folks involved. Needless to say, this isn't simple and the court of public opinion hasn't made it easy. I think folks understand that I can't share every detail of what occurred (as much as I'd personally love to). Although the full picture supports the actions Huntress has taken, I am not willing to sacrifice teammate privacy, law enforcement efficacy, or our integrity for the sake of reckless transparency. With that said, I'm looking to thread a needle to give as much clarity as I can through the lens of what actually happened—and just as importantly, what did not. What Did Not Happen We've conducted multiple investigations and we have found no evidence of illegal conduct or an insider threat, and consulted law enforcement who reached the same conclusion. When concerns were raised, we audited our systems thoroughly and found no evidence that: unauthorized access occurred, partner or customer data was disclosed, nor that source code or operational data was exposed. There was no "insider caught by the FBI". Based on the totality of the information we have gathered, we concluded that our partners and customers were not at risk then, and we have no reason to believe they are now. What Happened (and can be shared) Huntress permits threat researchers to occasionally engage with threat actors when it's beneficial for proactive R&D and/or to support active investigations. We are aware
Jun 30, 2026 · via huntress.com
The Arizona Department of Homeland Security and the Maricopa County Community College District recently opened the new Central Regional Security Operations Center (RSOC) located at the Glendale Community College.
The Central RSOC is the newest location in the state to help combat cybersecurity issues that come up in schools, towns, cities, tribal communities and more.
While providing online security for different partners and agencies, it’s also giving real-world experience to students who want to pursue a career in the cybersecurity industry that’s in need of more workers.
In the video player above, hear from interns and industry officials about how this new center will help with the workforce as well as the work they’re doing to help protect different communities.
Jun 30, 2026 · via abc15.com
Getty Images/WANAN YOSSINGKUM Who will shape the future of AI in the United States? Commentary Read more
Anthropic’s Mythos model found vulnerabilities in classified US government systems, official says Artificial Intelligence Read more
Jun 29, 2026 · via federalnewsnetwork.com