From the frontier AI front,

• The American Hospital Association News reports,
  • “Two critical vulnerabilities have been discovered in Hypertext Preprocessor, or PHP, a widely-used open-source scripted language for web development, its creators announced Oct. 21. The vulnerabilities reportedly impact a majority of the world’s websites. The two issues can allow unauthenticated remote code execution. The notice advises all PHP users to download updates immediately, as all PHP installations serving web requests are at risk. The National Security Agency advised organizations to take their systems offline if an immediate update is not feasible, and that the agency will soon issue specific guidance. An AHA Cybersecurity Advisory released Oct. 21 provides additional details and recommendations.”
    
• The Wall Street Journal explains why this is a frontier AI story.
  • “First, artificial intelligence heralded the Bugmageddon. Now it is triggering an all-hands-on-deck cyber emergency. 
  • “The creators of PHP, a technology that powers as much as 70% of the world’s websites, disclosed a critical security issue Wednesday that could be leveraged by hackers to take control of any system running its software. That would allow an attacker to steal information or launch a cyberattack on anyone visiting a hacked website.
  • “Security researchers found the vulnerability using AI. Now web administrators are scrambling to protect themselves from the vulnerability, which cutting-edge AI systems could quickly exploit.”
  • “Ryan Austin, a researcher at Calif, uncovered the flaws with help from GPT-5.6 Sol, one of OpenAI’s most advanced models, during an audit of PHP’s source code in August [2026]. OpenAI funded that and other audits, part of an effort by AI companies to help open-source projects address software vulnerabilities now being uncovered by artificial intelligence systems.
  • “The class of bugs had been hiding in PHP’s source code for years. A security researcher had flagged it as a problem area decades earlier, but the flaw had been considered too difficult to exploit and was largely forgotten, Duong said. 
  • “AI found a way. Austin figured out how to control the code that was running in the computer’s memory with as few as a few dozen requests to a server running PHP. On smaller systems, the hack takes just seconds. On more complex ones, it can take hours.” * * *
  • “PHP2Shell is “the most severe AI vulnerability that was ever found,” and one that was likely too complex to have been discovered without AI, said Ami Luttwak, Wiz’s co-founder. 
  • “Duong and the other researchers kept its existence a closely guarded secret while they worked on fixes, for fear that an AI-aided hacker might find and exploit it before they were done. 
  • “The patching effort was massive. There are close to 200 files being fixed and developers wrote over 10,000 lines of code during the process. 
  • “Eventually, a patch date was set: October 22. Then, the inevitable happened. News of the patch leaked out Tuesday, in a post to Hacker News, just under two days before the patch release.
  • “Dusch’s team decided to rush out the patch. On Tuesday, they warned the world to stand by and expect a critical security update on Wednesday [October 21].”

• Bleeping Computer and Cybersecurity Dive add,
  • [BC] “Security researchers at Amazon Threat Intelligence say attackers began targeting the PHP2Shell flaws four days before security patches were released on Wednesday [October 21]. * * *
  • “According to Amazon, the first exploitation attempts began on Saturday, October 17 (four days before the PHP2Shell patches were released), and attackers deployed multiple proof-of-concept PHP2Shell exploits hours after disclosure from VPN IP addresses.
  • “The exploit attempt originated from an IP address associated with a commercial anonymization network. While the vulnerability was under embargo, the targeting of a honeypot from an anonymized IP address suggests the activity was conducted by a threat actor, though it remains unclear how the actor acquired exploit information prior to disclosure,” said AWS chief information security officer CJ Moses.
  • “The actual payloads and choices in anonymization infrastructure varied, suggesting these were multiple actors. Technical analysis is ongoing, and it’s unclear if any of these attempts would have succeeded in real-world systems.”
  • “Moses added the PHP2Shell vulnerabilities are “now being broadly exploited in the wild,” though it’s unclear how many attacks are successful, and warned that “defenders should not assume the public disclosure date is the earliest date of exploitation.”
• and
  • [CD] “Security researchers warn that millions of websites remain at risk as exploitation attempts against two critical flaws in PHP have surged. 
  • “The vulnerabilities, tracked as CVE-2026-109618 and CVE-2026-109619, can enable an attacker to achieve remote-code execution. Each has a severity score of 9.2 out of 10. 
  • “Exploitation attempts have surged since the vulnerabilities were first disclosed, and researchers have uncovered threat activity from Potassium, a Mirai-derived botnet originally discovered in March. 
  • “Arctic Wolf said it observed an exploitation attempt that combined a “command-execution validation with a second-stage payload delivery.” 
  • “Two malicious binaries were identified within the payload hosting infrastructure, but Arctic Wolf said it did not find enough evidence to show successful downloading or execution. 
  • “Researchers from Sophos also observed threat activity linked to Potassium, but they were able to block an attack. An examination of the sample showed an updated and ‘significantly expanded’ version of the botnet variant, Sophos said.” 
    
• Here is a link to the October 23 alert from CISA about the PHP vulernability.
  • “CISA has updated this Alert [originally published October 21] to: (1) provide a SIGMA detection rule to help identify potentially suspicious activity; (2) incorporate partner resources; and (3) reinforce existing mitigation guidance.“
    
• Dark Reading offers expert guidance on mitigating the PHP vulnerability.
  • “Fixing the critical PHP flaws is not as simple as clicking “Update.” Here’s how to assess your risk, and what you can do to protect yourself.”

• The Wall Street Journal observes,
  • “The pace and volume of the AI-powered attacks are starting to overwhelm the incident-response firms that respond to cyber breaches. Pundits are referring to the month as “Black October,” and many in the field are predicting more to come.
  • “We expect that there’s going to be a massive number of compromise disclosures in the coming weeks and months,” said Steven Adair, president of the cybersecurity firm Volexity.
  • “Incident responders such as Adair say that AI tools are now being used by nation-state hackers from China and Russia as well as criminals looking to steal data and install ransomware. 
  • “One of the tools they are using is a Chinese open-source hacking tool called Artex, which uses artificial intelligence to automate the process of breaking into a computer network. 
  • “This week Microsoft said that “threat actors across multiple geographic regions” were using Artex to speed up attacks on hundreds of companies. Earlier this month, a cyberattack used Artex to steal data on about 68,000 people from South Korean banks.
  • “Hackers are using AI for reconnaissance, to analyze vast swaths of data they steal and even to rapidly build exploit code that can be used to break into networks, said Sherri Davidoff, founder of the security consulting firm LMG Security.”

From the cybersecurity and law enforcement front,

• Cyberscoop reports,
  • “A new bipartisan House bill would charge the federal government with developing formal guidance for defending AI models against distillation attacks.
  • “The AI Cyber Defense Act, shared exclusively with CyberScoop ahead of release, is sponsored by Reps. Pat Harrigan, R-N.C., and George Whitesides, D-Calif. It would direct the National Institute for Standards and Technology to develop voluntary industry standards and best practices for preventing distillation attacks from foreign entities from four countries: Russia, China, Iran and North Korea.
  • “In an interview, Harrigan told CyberScoop it is “incredibly expensive” for the U.S. AI industry to maintain its six-to-12 month lead over other countries, and distillation attacks give other parties an incentive to copy other models while keeping their own spending low.
  • “He said he believes countries such as China are making a strategic choice to hold back on full investment of their own AI industry, instead betting they can drift behind American companies as they perform the arduous and expensive research and development to advance the field.”

• Beckers Health IT relates,
  • “HHS will keep HIPAA civil monetary penalties at 2025 levels and skip its annual inflation adjustment for 2026, the department said in an Oct. 21 Federal Register notice.
  • “The freeze traces to last year’s federal government shutdown. The Bureau of Labor Statistics did not produce October 2025 consumer price index data during the lapse in appropriations from Oct. 1 to Nov. 12, 2025. Federal agencies need that month’s data to calculate the yearly penalty multiplier, so no 2026 multiplier exists.
  • “The Office of Management and Budget canceled the 2026 adjustments for federal agencies in an April 17 memo. HHS’ notice cites that guidance.” * * *
  • “The current HIPAA ceiling is about 46% higher than the $1.5 million cap set under the 2009 Health Information Technology for Economic and Clinical Health Act.”

• The Cybersecurity Hub Newsletter, writing in LinkedIn, informs us,
  • “The United States has ordered federal civilian agencies to remove DeepSeek technology from their information systems and exclude the Chinese artificial intelligence developer’s products from future procurement, widening restrictions on its role in government operations.
  • “A Department of Homeland Security notice made public on October 19 covers DeepSeek, affiliated companies and third-party products and services developed using DeepSeek implementations. The measure therefore reaches beyond the company’s chatbot to software supplied by other businesses that incorporates its technology.
  • “The official notice, scheduled for Federal Register publication invokes the Federal Acquisition Supply Chain Security Act of 2018. DHS says the decision followed a review of a Federal Acquisition Security Council recommendation and supporting evidence. Its stated scope concerns federal civilian procurement and information systems; it does not announce a nationwide prohibition on private use.”

• Tech Target reminds us,
  • “The U.S. government is moving closer to requiring covered critical infrastructure organizations to report significant cyber incidents within 72 hours and ransomware payments within 24 hours. But for operators of industrial control systems, the challenge will extend beyond deadline compliance; they must first determine whether an incident has occurred, establish what was compromised and decide whether the data driving physical processes can still be trusted.”

• Per Security Week,
  • “Effective and efficient use of complex agentic systems is important for all organizations, but recognizing, preventing and recovering from them going rogue could be an existential necessity.
  • “NIST, the NSA, and CISA recognize the increasing potential for AI agents to behave in an unexpected and potentially harmful manner, and has responded with a list of key practices for both preparing for and responding to such behavior. The report (Key Practices for Incident Readiness and Response in Super Intelligence Training and Evaluation) uses the term super intelligence (SI) rather than artificial intelligence (AI) in line with presidential executive order 14434.
  • “In the report, the ten key practices are grouped into five stages: prepare, detect, respond, recover, and report. For clarity, the report also includes three useful terminology definitions covering AI agents, AI control incident, and AI control near-miss.”

• The National Institute of Standards and Technology announced
  • “The initial public draft (ipd) of NIST Special Publication (SP) 800-100r1 (Revision 1), Information Security Handbook: A Guide for Executive Leadership, is now available for comment.” * * *
  • “The draft covers:
    • “An overview of key NIST frameworks and how they relate to one another
    • “Governance, including risk management fundamentals and enterprise risk management
    • “Topics organized under the CSF’s Identify, Protect, and Detect/Respond/Recover functions
    • “Continuous monitoring strategy and roles
  • “A public comment review period is open from October 12 to December 4, 2026. Commenters are encouraged to use the comment template and submit feedback to sp800-100-comments@nist.gov.”

From the cybersecurity breaches and vulnerabilities front,

• MedTech Dive reports,
  • “Quest Diagnostics is the latest medtech company to disclose that it was hit by a cyberattack.
  • “The testing services company identified unauthorized access to its Quanum Lab Services Manager system on Wednesday [October 21], according to a filing with the Securities and Exchange Commission. Quest’s Quanum LSM system is used by certain clients to place test orders as well as to receive and look up results. Quest paused access to the system upon discovering the incident.
  • “While Quest is still determining the full scope of the attack, it believes that an unauthorized party accessed certain patient records. Quest, in the filing, did not detail what patient records were accessed or how many records were involved.”

• HIPAA Journal lets us know that “Data breaches have been announced [recently] by UltraMobile Imaging in New York, Artistic Smiles in Colorado, Parkview Health in Indiana, and Cognitive Behavior Therapy Center in Virginia.

• Federal News Network fills us in about insider cybersecurity threats.
  • “Les Bernys, the executive director of the Department of Defense Cyber Crime Center, said the number of insider cyber incidents reported over the last six months far outpaced any other time in DC3 history.
  • “DC3 says more broadly, critical incident reports, which include insider threats, from defense companies have increased by over 50% in the last fiscal year. DC3 says it can’t give complete figures due to the anonymity of the DIB companies reporting.
  • Bernys said DC3 is seeing two kinds of insider threats.
  • “One type are former DoD employees who did malicious things to the network before they left or still have access to data after they left and were trying to exfiltrate data and, possibly, doing nefarious things with that data,” he said after speaking at the ACT-IAC Cybersecurity Summit on Oct. 14. “The second type of incident are companies are reporting that they have hired individuals who they thought were Americans but were North Korean. These incidents aren’t coming in huge numbers, but the increase is noticeable.”
  • “He added companies are struggling to better verify their applicants.”

• CISA added eight known exploited vulnerabilities to its catalog this week
  • October 20, 2026
    • CVE-2026-96652 Plex Media Server Server-Side Request Forgery (SSRF) Vulnerability
    • CVE-2026-96654 Plex Media Server Improper Neutralization of Encoded URI Schemes in a Web Page Vulnerability
    • CVE-2026-96655 Plex Media Server Server-Side Request Forgery (SSRF) Vulnerability
    • CVE-2026-96656 Plex Media Server External Control of File Name or Path Vulnerability  
    • CVE-2026-106931 EG Secure Solutions SiteGuard Server Edition OS Command Injection Vulnerability
    • CVE-2026-110936 Siemens Software Center Uncontrolled Search Path Element Vulnerability
      • The Plex Medica Service KVEs are PHP Shell vulnerabilities discussed in the frontier AI section above. Bleeping Computer also discusses the KVEs here.
      • Cybersecurity News discusses the EG Secure Solutions KVE here.
      • The Hacker News discusses the Siemens KVE here.
  • October 23, 2026
    • CVE-2026-114625 Arista VeloCloud Orchestrator SQL Injection Vulnerability
      • Security Week discusses this KVE here.
  • October 23, 2026
    • CVE-2026-110447 Collabora Online – Built-in CODE Server OS Command Injection Vulnerability 
      • GitHub discusses this KVE here.

• Bleeping Computer points out,
  • “GitLab has released security patches for a maximum-severity Regular Expression Compiler flaw that can let remote attackers run arbitrary code on vulnerable servers.
  • This security issue is tracked as CVE-2026-110264 and was reported by a security researcher using the ‘joaxcar‘ handle through GitLab’s HackerOne bug bounty program.
  • “CVE-2026-110264 stems from an integer overflow weakness in GitLab’s Regular Expression Compiler, a background engine that parses and validates regular expressions in configuration files.
  • “GitLab has fixed the vulnerability in GitLab Community Edition (CE) and Enterprise Edition (EE) versions 19.5.1, 19.4.3, 19.3.6, 19.2.9, 19.1.9, 19.0.10, and 18.11.13, and warned users to patch their instances immediately.
  • “GitLab has remediated an issue that, under certain conditions, could have allowed an unauthenticated user to execute arbitrary code on the GitLab server due to integer overflow when compiling a specially crafted regular expression,” the company said.”

• Cyberscoop tells us,
  • “Microsoft researchers attributed the recent supply-chain attack targeting ConnectWise customers’ Microsoft 365 environments to a China-state aligned threat actor it tracks as Storm-0940.
  • “As early as September 2026, Storm-0940 used the agentic AI penetration testing framework ARTEX to automate attacks targeting hundreds of organizations globally across a broad range of industries,” Microsoft Security Research wrote in a blog postWednesday. 
  • “The nation-state group, which has been active since 2021, targeted organizations in aerospace and defense, automotive, chemicals, energy, financial services, government, legal, manufacturing, media, military, semiconductors, software, technology, think tanks and non-profits. 
  • “Storm-0940 initiated a widespread, but ultimately unsuccessful, password-spraying campaign immediately preceding the compromise of ConnectWise’s Cloud Backup service. During a four-day stretch ending Oct. 14, Storm-0940 targeted 327 Microsoft 365 and Exchange Online accounts across eight organizations, researchers said.”

• Cybersecurity Dive informs us,
  • “An Indian company’s serious error in creating high-level encryption certificates has prompted major web browsers to block the certificates before hackers can use them to spoof websites.
  • “The issue involves intermediate certificates created by the Indian firm eMudhra, a company with the authority to issue certificates that are trusted by Google Chrome, Microsoft Edge, Mozilla Firefox and other browsers. Certificate issuers use intermediate certificates to “sign” individual website certificates that assure users of the sites’ legitimacy and security. But eMudhra recently failed to apply the proper procedures when creating eight of its intermediate certificates, producing keys that were shockingly weak.” * * *
  • “Leading web browsers responded by severing their trust relationships with the eight weak intermediate certificates. Google and Mozilla published blog posts detailing their responses to the discovery.
  • “Users should update their browsers and operating systems to the latest versions as soon as possible,” Dennis Jackson, a member of Mozilla’s cryptographic engineering team, wrote on an explainer page about the incident.
  • “There is no evidence that hackers generated any fraudulent website certificates by exploiting the weaknesses in eMudhra’s keys.”
• and
  • “Threat actors are deploying agentic AI in more sophisticated ways to automate aspects of their hacking campaigns, according to CrowdStrike.
  • “In 2026, CrowdStrike Intelligence has documented several incidents in which threat actors used AI not only to create tooling or artifacts, but to conduct agentic campaigns in which AI agents make decisions and execute commands,” the security firm said in a report published on Wednesday [October 21].

From the ransomware front,

• The American Hospital Association New reports,
  • “A joint advisory issued Oct. 20 by U.S. and international agencies warns of recent activity by the SafePay Ransomware group. The group has been active since 2024, attacking critical infrastructure organizations, including healthcare, across North America and around the world. The agencies said SafePay actors typically gain access to networks by stealing or guessing passwords, exploiting vulnerabilities in security systems and using phishing tactics. The actors then gain control of the network by creating new accounts or taking over existing ones and use legitimate system tools to move through the network to identify valuable files and disable security protections. Next, the actors encrypt the victim’s files and leave a ransom note demanding payment in cryptocurrency. The advisory includes recommended security measures to reduce the risk of SafePay ransomware attacks.”

• Cybersecurity Dive adds,
  • “Palo Alto Networks researchers warned that ransomware operators have stepped up exploitation of a high-severity flaw in the company’s GlobalProtect portal, months after it was first disclosed. 
  • “The authentication bypass vulnerability, tracked as CVE-2026-0257, allows an attacker to establish unauthorized VPN connections to the remote-access product. The flaw has a severity score of 7.8 out of 10. 
  • “Researchers warn that Qilin and INC ransomware groups are exploiting the vulnerability, according to a report by Unit 42, the threat intelligence and incident response arm of Palo Alto Networks. The company tracks those groups under the names Spikey Scorpius and Tarnished Scorpius, respectively.” 

• Info Security Magazine relates,
  • “Almost 3000 ransomware attacks took place between July and September 2026 and the figure marked a significant increase when compared with the same period last year.
  • “According to an analysis by Nord Security, the 2894 ransomware attacks which took place during the quarter represents a 50% increase when compared with the same period last year.
  • “It also represents a slight increase when compared with the previous quarter of this year, as the July to September period saw an 8.6% rise in ransomware attacks when compared with April to June.
  • “Published on October 21, the Nord Security report suggested that the significant rise in ransomware activity is a result of the impact of law enforcement takedowns on ransomware groups.
  • “Researchers suggested that the action against the likes of  Lockbit, BlackCat and RansomHubhelped to reduce ransomware activity during 2025.
  • “Now, the market is stabilizing, and the numbers we see today are merely the first stepping stone in ransomware’s new era,” said Vakaris Noreika, VP of product research and development at Nord Security.”

• Acronis takes “a closer look at a not-so-gentle ransomware-as-a- service operation.”
  • “The Gentlemen uses a double-extortion model. In addition to encrypting files, the ransomware can exfiltrate data from compromised systems. This puts additional pressure on victims to pay the ransom to restore access to their files and prevent the stolen data from being publicly leaked.”

• Tech Target offers a tip about
  • “How to tackle a ransomware readiness self-assessment, with template
    • “Ransomware readiness self-assessments are key to spotting vulnerabilities and preparing for third-party assessments. Use this template to evaluate your organization’s preparedness.”

From the cybersecurity business and defenses front,

• The Wall Street Journal reports,
  • “As businesses and consumers increasingly adopt artificial intelligence agents, 1Password is aiming to position itself to deliver solutions for growing challenges around security, access and trust.
  • “The Canadian late-stage cybersecurity startup has surpassed $500 million in annual recurring revenue, up from $400 million in November 2025, with a free cash flow margin in the double digits, the company said.
  • “The growth has been driven by an explosion in coding assistants and agents in enterprise settings, whose access to company secrets and systems needs to be closely managed, Chief Executive David Faugno said.
  • “The agents can actually take those secrets and do things that are unpredictable,” Faugno said in an interview. “You have to be very, very clear about what access you give it, and how you monitor and maybe even revoke that access, and humans have to stay in control.”

• Cybersecurity Dive relates,
  • “The size of the cybersecurity insurance market is growing quickly as more businesses try to guard against the damage from a major hacking incident.
  • “Total U.S. domestic cyber direct written business has nearly tripled since 2020, making cyber insurance one of the fastest-expanding segments in insurance,” the National Association of Insurance Commissioners said in a statement Wednesday [October 21] about its new report on the topic.
  • “Chubb, Beazley and St. Paul Travelers are the top three insurers by the value of their direct written premiums, according to the report.”

• Dark Reading adds,
  • “Cyber Insurance’s New Reality: Underwriting Isn’t Enough
    • “Rising AI, social engineering, and cloud risks are pushing carriers beyond risk transfer and deeper into cybersecurity itself.”

• Cyberscoop warns,
  • “Encryption isn’t enough to protect data wherever it goes 
    • “Once data is in use, encryption falls short. A tested approach that replaces sensitive values with realistic substitutes could make stolen information far less useful to hackers.”

• Per the Cybersecurity Hub Newsletter,
  • :Your SIEM detected the incident, Could you prove it six months later?
  • “That question comes up more often than it should. Usually during an audit.
  • “The incident was investigated. The threat was contained. The alert was closed. Six months later an auditor asks for the investigation record, the incident timeline, and the documentation showing what was reviewed and what was decided.
  • “The security controls worked. The paper trail did not.
  • “We put together the Managed SIEM Compliance Evidence Checklist for NIS2 and DORA to help teams find those gaps before a review does.”

• Here’s a link to Dark Reading’s CISO Corner.