IEEE Cipher --- Items from security-related news (E190)





Prior news summaries from Cipher




  • AI Creates Teams of Hacking Agents
    Chinese Hackers Use Anthropic's AI to Launch Automated Cyber Espionage Campaign

    Publisher:The Hacker News
    Date: November 14, 2026
    By: Ravie Lakshmanan

    Summary:
    Anthropic may have disagreements with the US Government, but last September it helped to reveal the urgency of protecting computer systems against agentic AI systems. The company discovered that its AI system, Claude, acting as several different "agents", carried out a sophisticated cyberattack against multiple targets last September. The human attackers benefited from the ability of the software agents to act quickly and cooperatively in carrying out an attack. Human intervention occurred only at strategic junctures.

    "Threat actors can now use agentic AI systems to do the work of entire teams of experienced hackers with the right set up, analyzing target systems, producing exploit code, and scanning vast datasets of stolen information more efficiently than any human operator. Less experienced and less resourced groups can now potentially perform large-scale attacks of this nature."

    Anthropic has taken steps to detect and prevent similar exploits, but it does appear that the genie has escaped the bottle.


  • OpenClaw Bites the Hand that Serves it
    ClawJacked Flaw Lets Malicious Sites Hijack Local OpenClaw AI Agents via WebSocket

    Publisher: The Hacker News
    Date: Feb 28, 2026
    By: Ravie Lakshmanan

    Summary:
    The OpenClaw AI Agent included a vulnerability that could be used by an attacker running a website that might be visited by someone running an OpenClaw system. Because OpenClaw uses a local websocket server protected only by a password, the remote system can use high-speed password guessing to gain a trusted connection back to OpenClaw user's local machine. The problem was patched within a day of notification to OpenClaw.

    Protecting something as mundane as a network connection might have been far below the radar of OpenClaw engineers because they have put effort into protecting against attacks that might be enabled by their core capability: user provided instructions for AI agents trained on the user's personal, local, activities. Those instructions might contain malicious code, and OpenClaw has tried to assure that they detect such code through tools such as VirusTotal


  • The Citizen Hacker
    Trump Calls On Private Companies to Take On a Bigger Role in Cyber
    In his National Cybersecurity Strategy, President Trump sought an expanded role for private firms in cyberwarfare. He did not take on China or Russia in the document.
    Publisher: The New York Times
    Date: March 6, 2026
    By: Adam Sella

    Summary:
    The White House issued a brief statement with the National Cybersecurity Strategy. The seven page document seems to encourage private companies to conduct cyberattacks against American enemies, something that is not currently allowed. "We will unleash the private sector by creating incentives to identify and disrupt adversary networks and scale our national capabilities."

    Other notable excerpts: "We will promote the adoption of post-quantum cryptography ..."
    This is an expensive undertaking and worthwhile only if quantum computing succeeds. Another part of the report emphasizes commitment to quantum computing development.
    "We will work to adopt AI-powered cybersecurity solutions ..."
    Cyber criminals are working now to adopt AI-powered cyberattack capabilities. Presumably the call for acting against our cybersecurity enemies will include similar US activities.


  • Medical Tech Company Victim of Iranian Cyberattack
    Iran appears to have conducted a significant cyberattack against a U.S. company, a first since the war started
    The company, Stryker, said a cyberattack disrupted its "Microsoft environment."
    Publisher: NBC News
    Date: March 11, 2026
    By: Kevin Collier

    Summary:

    An Iranian hacker group, Handala, claimed responsibility for a cyberattack on a medical tech company, Stryker. The company suffered disruption of its employee phones and loss of data.

    This was the first report of a cyberattack during the US-Iran war. Other cyber activity by Iranian groups had been for intel gathering.


  • Department of State Protects the Homeland
    State Department launches effort to counter cyberattacks, AI risks from Iran, others
    Officials detailed the Bureau of Emerging Threats exclusively to ABC News.
    Publisher: ABC News
    Date: March 23, 2026
    By: Shannon K. Kingston

    Summary:
    A year ago Secretary of State Marco Rubio announced an overhaul to the department that would include the Bureau of Emerging Threats. Recently, the structure of new Bureau was described as having five offices: the Office of Cybersecurity, the Office of Critical Infrastructure Security, the Office of Disruptive Technology, the Office of Space Security and the Office of Threat Assessment. A spokesman for the Department said that the Bureau would address both current and future challenges posed by disruptive technology.


  • AppArmor Hands Over Unauthenticated Policy Data
    Nine CrackArmor Flaws in Linux AppArmor Enable Root Escalation, Bypass Container Isolation
    Publisher: The Hacker News
    Date: Mar 13, 2026
    By: Ravie Lakshmanan

    Summary:

    The Qualys Threat Research Unit (TRU) said that they identified flaws in the Linux security module AppArmor. The module helps isolate applications from each other by enforcing a mandatory access control (MAC) policy. The flaws allow privilege escalation by unauthorized parties.

    "The problem affects all Linux kernels since version 4.11 on any distribution that integrates AppArmor. With more than 12.6 million enterprise Linux instances operating with AppArmor enabled by default in several major distributions, such as Ubuntu, Debian, and SUSE, immediate kernel patching is advised to mitigate these vulnerabilities."

    The flaws are instances of "... the 'Confused deputy' problem, in which an attacker misuses the authority of one victim (the 'confused deputy') to use that victim's legitimate (restricted) capabilities to target another victim." In this case, the AppArmor module is the Confused Deputy, forwarding attacker-crafted data that cause security profiles to be bypassed.


  • America First in Routers
    Everything You Need to Know About the Foreign-Made Router Ban in the US
    The FCC just banned the sale of new consumer-grade Wi-Fi routers manufactured outside the US. Here's what it means for you.
    Publisher:
    Date: Mar 24, 2026
    By: Simon Hill

    Summary:

    Sales of Internet routers, the kind that ordinary consumers have in their homes, will be restricted by new rules announced by the Federal Communications Commission (FCC). Routers made outside the US will only be sold to US consumers if the manufacturers applies for and receivers a waiver to the "no foreign routers" policy. This does not affect routers that have already been purchased. The policy has a FAQ page.

    Most routers are manufactured outside the US or have foreign parts, so the policy might result in a shortage of consumer routers, at least until manufacturers can adapt to it.

    Concern about security and routers is not new, but the reasons for imposing the ban at this time are not known. There has been talk of banning TP-Link routers because of the company's link to China. They have been accused of allowing Chinese government access to their routers and of flooding the market with underpriced devices. Their manufacturing is in Viet Nam.