SEC595: Applied Data Science and AI/Machine Learning for Cybersecurity Professionals


Experience SANS training through course previews.
Learn MoreLet us help.
Contact usBecome a member for instant access to our free resources.
Sign UpWe're here to help.
Contact UsThe US Cybersecurity and Infrastructure Security Agency (CISA) has published a Binding Operational Directive (BOD) instructing federal civilian executive branch (FCEB) agencies to remove all end-of-support (EoS) edge devices from their networks. CISA notes that "edge devices are attractive targets due to their extensive reach into an organization's network and integrations with identity management systems." The BOD is aligned with an Office of Management and Budget document, "Managing Information as a Strategic Resource," which requires agencies to phase out EoS systems and components. Agencies are advised to update IT lifecycle management practices, to include identifying software and hardware approaching EoS and planning accordingly for decommissioning and replacement. The BOD requires FCEB agencies to immediately update supported devices running EoS software or firmware to supported versions unless doing so would "adversely impact mission critical functionality." Within three months, agencies must provide CISA with an inventory of EoS edge devices; CISA has provided a list of edge devices that will soon reach or have already reached EoS. Within one year, agencies must decommission all EoS edge devices from the list with EoS dates of 12 or fewer months, report the decommissioning to CISA, and provide CISA with an inventory of edge devices that have EoS dates falling within the following 12 months. There are also 18- and 24-month requirements that help agencies establish continuous processes for managing EoS edge devices.

Managing system lifecycle is critical. Prioritize perimeter and boundary control devices. We've seen too many exploits of past-service-life devices, particularly routers, VPNs, and firewalls, FCEB or otherwise. This new directive, BOD-26-02, gives agencies a generous timeline to decommission/replace all EOS devices; given the current threat environment, consider giving that much higher priority. Of note, FCEB agencies are bound by multiple guidelines which require this including NIST 800-207, Zero Trust Architecture, and NIST 800-39, Risk Management Framework. Interestingly, NIST 800-53 does include SDLC in control SA-3, but the requirement for technology refresh, SA-3(3), is not part of any baseline. If you're uncertain about ongoing need to keep current, consider adding SA-3(3) to your security plans. Make sure technology refresh is in the budget and included in future submissions.
A prudent measure by CISA. Three months to inventory and a year to actually replace the EoS product seems a tad long, but then it apparently hasn't been managed in the past. They should extend this to include all internet-facing, end of life products.

While this is certainly good security practice, it may also be very expensive. The Directive is likely to find agencies unprepared, not to say unable.
CISA
Nextgov/FCW
CyberScoop
Gov Infosecurity
The Record
Researchers are warning that a threat group linked to Russian military intelligence has been conducting cyberespionage targeting primarily "maritime and transport organizations across Poland, Slovenia, Turkey, Greece, the UAE, and Ukraine," starting to exploit a vulnerability in Microsoft Office within 24 hours of the company's January 26, 2026 emergency security advisory addressing the flaw. CVE-2026-21509, CVSS score 7.8, allows an unauthorized local attacker to bypass Object Linking and Embedding (OLE) mitigations when a targeted user opens a malicious Office file, due to "reliance on untrusted inputs in a security decision." The campaign took the form of a "concentrated 72-hour spear-phishing campaign," exploiting specific "geopolitically charged narratives" in the lures, mentioning weapons shipments, military training programs, diplomatic consultations, and emergency weather bulletins. Once the target opens the malicious Office file, a crafted OLE object executes, retrieving a payload that begins an infection chain resulting in the installation of backdoor malware. Trellix notes this threat actor's "heavy abuse of legitimate cloud services," among other tactics, techniques, and procedures (TTPs) allowing the researchers to attribute the attack to APT28 with high confidence. The researchers provide indicators of compromise (IoCs), also noting the utility of MITRE ATT&CK mapping for threat hunting. "Organizations are urged to apply the latest Office patches (including the emergency fix for CVE-2026-21509) and implement Microsoft’s recommended registry hardening that blocks this OLE exploit path."

Our old friend untrusted input is back. And while Microsoft has taken steps to release fixes and provide guidance, the ball is in our court. Remember those emergency Office updates we talked about the other day? Make sure they got pushed out, and if you're on Office 2016 or 2019, you need the registry keys. If you're on Office 365, make sure you've restarted the desktop office tools.
There was reason why MSFT published the emergency security advisory. One shouldn’t wait for linkage to a threat actor before taking action. We also now have confirmation that attackers can operate within a 24-hour clock to exploit a vulnerability. That’s a good KPI for security professionals to use to enforce patch management of critical vulnerabilities.
Trellix
Zscaler
The Record
Ars Technica
Dark Reading
The Hacker News
The Register
On February 3, 2026, Microsoft's Azure storage ended support for Transport Layer Security (TLS) 1.0 and 1.1, making TLS 1.2 the new minimum TLS version. TLS 1.0 was introduced in 1999, and TLS 1.1 was introduced in 2006; both were deprecated in 2021. TLS 1.2 was introduced in 2008, and version 1.3 was published in 2018. Microsoft has held onto TLS 1.0 and 1.1 for this long due in large part to backward compatibility. In October 2024, Microsoft cautioned users that "all clients connecting to Azure storage need to send and receive data by using TLS 1.2, [and] otherwise will not be able to connect to storage using SSL connections." In a separate story, Microsoft has also announced that Exchange Web Services (EWS) will reach end-of-service (EoS) starting later this year, with shut down slates for April 2027. Starting in October 2026, Microsoft will block Exchange Online EWS by default; admins will be able to temporarily maintain access through application allow lists. Microsoft says of the decision, "EWS was built nearly 20 years ago, and while it served the ecosystem well, it no longer aligns with today’s security, scale, or reliability requirements."

Take a look and make sure you've really eliminated all your TLS 1.0 and 1.1 use. Verify it's really long gone, and that you're well on the way to using TLS 1.3, as TLS 1.2 was supposed to go EOL August 31, 2025. You don't want to find out you missed something important because something stopped working, as it'll never be something that doesn't have to be immediately addressed.
Finally. It’s crazy that MSFT kept TLS 1.0 and 1.1 in service five years after formal deprecation, but the reality is that customers get a vote. I guess in the intervening two years enough customers migrated to make it manageable shutting down the deprecated protocols (i.e., less blowback on the Softies).
The Register
Microsoft
Microsoft
BleepingComputer
Microsoft
Palo Alto Networks Unit 42 has published an investigation report warning of cyberespionage and data theft attacks targeting global governments and critical infrastructure in the past year, dubbed "Shadow Campaigns." The attacks have previously taken the form of phishing campaigns, which trick the target into downloading a ZIP file containing an executable that ultimately installs a Cobalt Strike or VShell payload. The attackers focus on moving laterally and maintaining persistence, employing a wide variety of tools, exploitation kits, and proof-of-concept exploits, including web shells, tunnels, and a previously unseen Linux kernel rootkit. Unit 42 tracks the threat actor as TGR-STA-1030, a state-aligned group based in Asia, active for at least two years. TGR-STA-1030 has been observed conducting reconnaissance scans on 155 countries and successfully breaching at least 70 organizations across 37 countries in Africa, the Americas, Asia, Europe, and Oceania, targeting "ministries and departments of interior, foreign affairs, finance, trade, economy, immigration, mining, justice and energy." Pete Renals, Unit 42's Director of National Security Programs, stated to news sources that "this is probably the most widespread and significant compromise of global government infrastructure by a state-sponsored group since SolarWinds." Unit 42 has contacted affected organizations and is continuing to monitor the threat actor's ongoing activity; the report provides indicators of compromise (IoCs) for threat hunters.

The Unit 42 report provides not only analysis of the exploit but also country-by-country analysis of TGR-STA-1030's activities. CISA is also tracking this threat actor. The report lists IoCs which include a hash for attempted exploits of CVE-2019-11580, the old Atlassian Crowd development plugin flaw. Not a bad time to verify your Atlassian environment is up to snuff. Mitigations for the attack include URL filtering and DNS security, network protections, and watching for abnormal behavior. Maybe time to visit your MDR team to get their take on protecting from this sort of attack.
Unit42
The Record
The Register
Axios
Romanian pipeline operator Conpet is investigating a cyberattack that disrupted the company's corporate systems earlier this week. The incident has not disrupted Conpet's operational technology, although the company website is currently unavailable. La Sapienza, a university in Rome, Italy, has experienced technical disruptions due to a cyberattack. The university is working with Italian CSIRT, experts from Agenzia per la Cybersicurezza Nazionale (ACN), and the Polizia Postale to restore the school's systems from backups.

Incidents like this with the Romanian pipeline operator show how often “corporate system” outages are treated as separate from operations, even though they directly affect situational awareness and coordination. A pipeline can continue to flow, but when communications, websites, and business systems are down, operators and partners lose visibility and trust.

La Sapienza in Rome has about 120,000 students and reportedly took down systems out of caution to contain the threat, as well as setting up "infopoints" on campus to provide information to students. University exams are continuing as normal, but students must sign up for them directly with professors rather than online. A previously unknown group, Femwar02, is reported to be behind the La Sapienza attack, using the BabLock (aka Rorschach) ransomware. Based on the recovery actions being taken, I expect the university will not pay any ransom demand.

When a pipeline operator in the US was attacked, OT was not directly attacked but service was disrupted because the billing system was taken down.
On Wednesday, February 4, Cisco published five security advisories to address vulnerabilities in multiple products. The flaws include a high-severity arbitrary file upload vulnerability in Cisco Meeting Management (CVE-2026-20098); a high-severity denial-of-service vulnerability in Cisco TelePresence Collaboration Endpoint Software and RoomOS Software (CVE-2026-20119); a medium-severity file bypass vulnerability in Cisco Secure Web Appliance (CVE-2026-20056); a medium-severity stored cross-site scripting vulnerability in Cisco Prime Infrastructure (CVE-2026-20111); and a medium-severity open redirect vulnerability in Cisco Evolved Programmable Network Manager and Cisco Prime Infrastructure (CVE-2026-20123). Also on Wednesday, February 4, F5 published its Quarterly Security Notification, which addresses three medium severity and two low-severity vulnerabilities in the company's BIG-IP and NGINX products.

The updates from F5 include a patch for your WAF/ASM which can be used to cause the bd process to crash, resulting in a DOS to your application/website. The NGINX flaw could be used to inject fake responses which get to clients, and the container ingress services flaw could be used to read container secrets. Beyond application of the updates, make sure that your management and control plane interfaces are protected from unauthorized access.
In a recent blog post, GreyNoise Senior Director of Security Research and Detection Engineering Glenn Thorpe writes that last year, the US Cybersecurity and Infrastructure Security Agency (CISA) changed the "Known To Be Used in Ransomware Campaigns" status of 59 CVEs in the Known Exploited Vulnerabilities (KEV) catalog from Unknown to Known, indicating what Thorpe calls "a material change in your risk posture." However, CISA does not issue notifications when a CVE's ransomware use status changes. Research indicates that KEV CVEs designated as known to be used in ransomware campaigns are patched or mitigated 2.5 times faster than those without the designation. Of the 59 CVEs flipped in 2025, 16 involved Microsoft vulnerabilities, six Ivanti vulnerabilities, five Fortinet vulnerabilities, and three vulnerabilities each for Palo Alto Networks and Zimbra. The time between the appearance of the CVE in KEV and the ransomware status change ranges from one day to 1,353 days. In response, GreyNoise has created an RSS feed that alerts subscribers to changes in KEV CVE ransomware status changes.

You're likely ingesting KEV and CVE data into your threat analysis processes, so add the GreyNoise RSS feed to catch these state changes. Stack the deck in your favor. Consider filing a polite enhancement request with CISA as well to issue notifications when this state changes.
Organizations use the KEV catalog to help set prioritization. If a vulnerability makes it onto the KEV list, it demands immediate attention regardless of later change in use status. The recent reporting on the MSFT Office OLE vulnerability and time to exploit highlights what defenders are up against. The motto, “patch early, patch often” seems prudent here.
Newsletter platform Substack is emailing users to disclose a data breach that took place in October 2025 but was not discovered until a system issue was identified on February 3, 2026. Substack states that they have fixed the problem that allowed an unauthorized third party to access user data and are conducting an investigation while working on system improvements. The affected data include email addresses, phone numbers, and internal metadata; Substack encourages users to be wary of suspicious texts and emails. The nature and scope of the attack have not yet been disclosed.

Substack reports that while email addresses and phone numbers were exfiltrated, passwords, financial information and credit card numbers were not. If you're a Substack user, watch for attempts to phish you with that account information. Maybe double check that email and phone number combination is included in your ID monitoring.
By now, I think everyone is on guard for suspicious texts and emails. Let’s add voice messages to the mix as well. The Substack notification doesn’t change that reality, but a reminder is helpful. That said, I do like the unusually candid language used by the CEO in the email message. The next step is to share more details on what happened and how you’re making lasting changes in your cybersecurity program.
BleepingComputer
TechCrunch
The Record
The Register
SecurityWeek
Authorities in Poland have arrested a suspect in connection with a series of distributed denial-of-service (DDoS) attacks on "numerous popular websites" around the world, according to the Central Bureau for Combating Cybercrime (CBZC). Authorities seized computer equipment from the suspect's residence. The unnamed individual is facing charges of disrupting an IT system and obtaining a computer program used for such disruption; if convicted, he could face up to five years in prison. The investigation into the attacks is ongoing and further arrests have not been ruled out.

Poland's CBZC is putting attackers on notice: they are going to find them and consequences will ensue, particularly for those involved in DDoS. They are also participating in larger Europol operations, such as Operation PowerOFF and Operation Eastwood, and are seeing a 30 percent increase in cybercrime charges and detentions in 2025. CBZC has nearly doubled in size since 2024, having added 1000 staff in the past year. This is not a good time to be operating cybercrime out of Poland.
The Register
Help Net Security
SANS Internet Storm Center StormCast Friday, February 6, 2026
Broken Phishing; n8n vulnerability; Android Update; Watchguard Firebox LDAP Injection
https://isc.sans.edu/podcastdetail/9798
Broken Phishing URLs
https://isc.sans.edu/diary/Broken+Phishing+URLs/32686/
n8n command injection vulnerability
https://github.com/n8n-io/n8n/security/advisories/GHSA-6cqr-8cfr-67f8
Android February Update
https://source.android.com/docs/security/bulletin/pixel/2026/2026-02-01?hl=en
Watchguard Firebox LDAP Injection
https://www.watchguard.com/wgrd-psirt/advisory/wgsa-2026-00001
SANS Internet Storm Center StormCast Thursday, February 5, 2026
Malicious Scripts; Synectix Vuln; Google Chrome; Google Looker
https://isc.sans.edu/podcastdetail/9796
Malicious Script Delivering More Maliciousness
https://isc.sans.edu/diary/Malicious+Script+Delivering+More+Maliciousness/32682
Synectix LAN 232 TRIO Unauthenticated Web Admin CVE-2026-1633
https://www.cisa.gov/news-events/ics-advisories/icsa-26-034-04
Google Chrome Patches
https://chromereleases.googleblog.com/2026/02/stable-channel-update-for-desktop.html
LookOut: Discovering RCE and Internal Access on Looker (Google Cloud & On-Prem)
https://www.tenable.com/blog/google-looker-vulnerabilities-rce-internal-access-lookout
SANS Internet Storm Center StormCast Wednesday, February 4, 2026
Detecting OpenClaw; Synology telnetd Patch; More GlassWorm
https://isc.sans.edu/podcastdetail/9794
Detecting and Monitoring OpenClaw (clawdbot, moltbot)
https://isc.sans.edu/diary/Detecting+and+Monitoring+OpenClaw+clawdbot+moltbot/32678
Synology telnetd Patch
https://www.synology.com/en-us/releaseNote/DSM
GlassWorm Loader Hits Open VSX via Developer Account Compromise
https://socket.dev/blog/glassworm-loader-hits-open-vsx-via-suspected-developer-account-compromise
Catch up on recent editions of NewsBites or browse our full archive of expert-curated cybersecurity news.
Browse ArchiveCISO Best Practices Cheat Sheet. Whether you’re inheriting a cloud program, scaling to multi-cloud maturity, or aligning with board priorities, this cheat sheet helps you cut through the noise, focus on measurable outcomes, and lead with clarity - all with practical frameworks and 90-day actionable steps.
Webcast | Stop SIEM Cost Bloat: Expand Visibility Without Expanding Your Budget.
Free Event | SANS 2026 Winter Cyber Solutions Fest | February 11-12, 2026. Join us for this two-day event focused on finance, healthcare and critical infrastructure. Register for one topic or for all three.
Take the 2026 SANS Detection Engineering Survey. Share your knowledge with the community and be entered to win a $250 Amazon gift card.