← Terug naar nieuwsbrief

AI & Security Nieuws – 8 December 2025

Nieuwsbrief

By Djimit* een overzicht voor AI cloud- en security professionals*

Hoofdtrend

De dominante trend is verscherpte machtsconcentratie rond frontier-modellen gecombineerd met een groeiende markt voor evaluatie- en governance-oplossingen: terwijl OpenAI versnelt richting GPT‑5.2 onder concurrentiedruk van Google’s Gemini 3, positioneren leveranciers als Seekr zich expliciet als compliance- en risicolaag bovenop deze modellen. Parallel groeit de aandacht voor bestuurlijke implicaties van deze concentratie, waarbij analyses benadrukken dat een klein cluster bedrijven de infrastructuur, data en talent controleert die de ontwikkelcurve van AI bepaalt.[2][3][4][1]

1. OpenAI versnelt GPT‑5.2 na ‘code red’

2. SeekrGuard: commerciële evaluatielaag voor AI‑risico’s

3. Info‑Tech: waarom AI‑investeringen blijven steken

4. Strategische analyse: leiderschap en machtsconcentratie in AI

5. Domeinspecifiek: gedragsredeneren voor autonome voertuigen

6. Data & monitoring: Epoch AI‑updates over modeltrends

Insight of the Day

De combinatie van een versneld GPT‑5.2‑release‑traject en de lancering van commerciële evaluatie‑ecosystemen als SeekrGuard laat zien dat de “frontier race” en de “evaluation/governance‑laag” nu parallel volwassen worden: snelheid van modelreleases zonder ingebouwde, continue evaluatie‑infrastructuur wordt strategisch onhoudbaar voor serieuze gebruikers.[1][2][3]

Bronnen[1] OpenAI rushes out new AI model after ‘code red’ declared https://ca.news.yahoo.com/openai-rushes-ai-model-code-105822611.html[2] Seekr Launches AI Evaluation Product to Enable Compliance with … https://www.morningstar.com/news/pr-newswire/20251208ny41045/seekr-launches-ai-evaluation-product-to-enable-compliance-with-the-presidents-ai-action-plan[3] This changes everything: Business leadership in the age of AI https://www.reuters.com/sustainability/boards-policy-regulation/this-changes-everything-business-leadership-age-ai–ecmii-2025-12-08/[4] OpenAI could release GPT‑5.2 this week to challenge … https://www.windowscentral.com/artificial-intelligence/openai-chatgpt/openai-is-racing-to-give-chatgpt-a-flashy-upgrade[5] Why AI Investments Stall: New Guidance on Data Operating Models … https://www.prnewswire.com/news-releases/why-ai-investments-stall-new-guidance-on-data-operating-models-published-by-info-tech-research-group-302635712.html[6] Can AI read humans’ minds? A new model shows it’s … https://stories.tamu.edu/news/2025/12/08/can-ai-read-humans-minds-a-new-model-shows-its-shockingly-good-at-it/[7] AI’s Breakthroughs for the Week – December 1 – 7, 2025 https://www.kinglyclark.com/blog-post/ai-weekly-wrap-ais-breakthroughs-for-the-week—december-1—7-2025[8] Epoch AI https://epoch.ai[9] AI Weekly Digest December 8 2025 – LinkedIn https://www.linkedin.com/pulse/ai-weekly-digest-december-8-2025-pa-media-qqmoe[10] EU Digital Omnibus: The European Commission Proposes Important … https://www.sidley.com/en/insights/newsupdates/2025/12/eu-digital-omnibus-the-european-commission-proposes-important-changes-to-the-eus-digital-rulebook[11] URI’s use of artificial intelligence in research, teaching, and … https://www.uri.edu/news/2025/12/uris-use-of-artificial-intelligence-in-research-teaching-and-innovation-on-display-at-discovering-ai-uri-day/[12] JAX receives an up to $30M contract award to revolutionize … https://www.jax.org/news-and-insights/2025/december/jax-receives-an-up-to-30m-contract-award-to-revolutionize-drug-safety-testing-with-ai-powered-virtual-hearts[13] Technology News Today – The Latest in Tech, AI & Startup News … https://techstartups.com/2025/12/08/technology-news-today-the-latest-in-tech-ai-startup-news-december-8-2025/[14] OpenAI plans to release a new ‘open’ AI language model in … https://techcrunch.com/2025/03/31/openai-plans-to-release-a-new-open-language-model-in-the-coming-months/[15] The European Journal on Artificial Intelligence https://journals.sagepub.com/home/eai[16] What’s New in Artificial Intelligence – December 2025 – AI Toolr https://www.ai-toolr.com/en/blog/what%E2%80%99s-new-in-artificial-intelligence-%E2%80%93-december-2025[17] OpenAI’s long-awaited GPT-5 model nears release https://www.reuters.com/business/retail-consumer/openais-long-awaited-gpt-5-model-nears-release-2025-08-06/[18] Novel AI technique able to distinguish between … https://www.yorku.ca/news/2025/12/08/novel-ai-technique-able-to-distinguish-between-progressive-brain-tumours-and-radiation-necrosis-york-u-study-finds/[19] OpenAI rushes out new AI model in ‘code red’ response to … https://www.aol.com/articles/openai-rushes-ai-model-code-105822130.html[20] OpenAI Declares “Code Red” as Google Catches Up in AI … https://llmrefs.com/blog/openai-code-red-chatgpt

Kritieke Incidenten

Agfa-Gevaert concludeerde op 8 december 2025 een onderzoek naar een cybersecurity-incident waarbij de ransomwaregroep Everest claimde toegang te hebben verkregen tot een bestandsserver. Het incident beperkte zich tot oudere, niet-gevoelige data zonder impact op kritieke systemen of persoonlijke informatie, en alle operaties bleven volledig operationeel.[1][2]

Zero-Day en Vulnerabilities

De kritieke React2Shell-kwetsbaarheid (CVE-2025-55182, CVSS 10.0) in React Server Components wordt actief geëxploiteerd in China-gekoppeld APT-verkeer, met publieke proof-of-concept exploits beschikbaar sinds 4 december 2025. Deze unauthenticated RCE treft frameworks als Next.js en vereist dringende patching buiten reguliere cycli.[3][4]

Google patchte op 7-8 december 2025 twee Android zero-days (CVE-2025-48633 informatielek, CVE-2025-48572 privilege-escalatie) plus CVE-2025-48631 (kritieke Framework DoS), beide onder limited targeted exploitation.[5]

APT-Campagnes

OceanLotus (APT32) lanceerde op 8 december 2025 een geavanceerde cyberespionage-campagne tegen de Xinchuang-ecosysteem in China, gericht op geavanceerde persistent threats met sophisticated malware.[6]

Bronnen[1] Agfa concludes investigation into cybersecurity incident https://www.agfa.com/corporate/news-item/agfa-concludes-investigation-into-cybersecurity-incident/[2] Agfa Gevaert Concludes Investigation Into Cybersecurity Incident https://www.tradingview.com/news/reuters.com,2025:newsml_FWN3XC0DF:0-agfa-gevaert-concludes-investigation-into-cybersecurity-incident/[3] H-ISAC TLP Green: Daily Cyber Headlines – December 8, 2025 | AHA https://www.aha.org/h-isac-green-reports/2025-12-08-h-isac-tlp-green-daily-cyber-headlines-december-8-2025[4] React2Shell (CVE-2025-55182) – Critical unauthenticated RCE … https://www.rapid7.com/blog/post/etr-react2shell-cve-2025-55182-critical-unauthenticated-rce-affecting-react-server-components/[5] Google Patches 107 Android Vulnerabilities, Including Zero-Days https://www.gopher.security/news/google-patches-107-android-vulnerabilities-including-zero-days[6] OceanLotus Targets Xinchuang Ecosystem with Sophisticated … https://gbhackers.com/xinchuang-ecosystem/[7] Two Android 0-day bugs patched, plus 105 more fixes – The Register https://www.theregister.com/2025/12/02/android_0_days/[8] Chinese APT Target Royal Thai Police in Malware Campaign https://www.darktrace.com/blog/chinese-apt-target-royal-thai-police-in-malware-campaign[9] November 2025: Major Cyber Attacks, Ransomware … https://www.cm-alliance.com/cybersecurity-blog/november-2025-major-cyber-attacks-ransomware-attacks-data-breaches[10] 8th December – Threat Intelligence Report – Check Point Research https://research.checkpoint.com/2025/8th-december-threat-intelligence-report/[11] APT36 Targets Indian Government with Golang-Based DeskRAT … https://thehackernews.com/2025/10/apt36-targets-indian-government-with.html[12] 27 Biggest Data Breaches Globally (+ Lessons) 2025 https://www.huntress.com/blog/biggest-data-breaches[13] The Most Recent Data Breaches in 2025 – Breachsense https://www.breachsense.com/breaches/[14] Google Fixes 107 Android Vulnerabilities, Including Two Actively … https://www.linkedin.com/pulse/google-fixes-107-android-vulnerabilities-including-tmcce[15] Kaspersky predicts AI backdoors, hacktivism in APT attacks https://backendnews.net/forecast-2025-kaspersky-predicts-ai-backdoors-hacktivism-in-apt-attacks/[16] The top data breaches of 2025 so far https://nordlayer.com/blog/data-breaches-in-2025/[17] Oracle EBS zero-day used by Clop to breach Barts Health NHS https://securityaffairs.com/185447/data-breach/oracle-ebs-zero-day-used-by-clop-to-breach-barts-health-nhs.html[18] Risky Bulletin: APTs go after the React2Shell vulnerability within hours https://news.risky.biz/risky-bulletin-apts-go-after-the-react2shell-vulnerability-within-hours/[19] FTC Takes Action Against Education Technology Provider for Failing … https://www.ftc.gov/news-events/news/press-releases/2025/12/ftc-takes-action-against-education-technology-provider-failing-secure-students-personal-data[20] The Hacker News | #1 Trusted Source for Cybersecurity News https://thehackernews.com[21] December 2025 Android Security Bulletin: Two Zero-Day Flaws … https://socradar.io/december-2025-android-security-bulletin/[22] 2025 – Data Breaches Digest https://www.dbdigest.com/2025/[23] [PDF] Cybersecuritybeeld Nederland 2025 https://www.nctv.nl/site/binaries/site-content/collections/documents/2025/11/26/cybersecuritybeeld-nederland-2025/cybersecuritybeeld-nederland-2025.pdf[24] M&S, Co-op cyberattackers duped IT help desks into resetting passwords, says report https://www.reuters.com/business/retail-consumer/ms-co-op-cyberattackers-duped-it-help-desks-into-resetting-passwords-says-report-2025-05-06/[25] Everest claimt cyberaanval Agfa – Printmedianieuws.nl https://www.printmedianieuws.nl/2025/11/everest-claimt-cyberaanval-agfa/[26] Security.NL https://www.security.nl[27] Cyber Briefing: 2025.12.08 https://www.linkedin.com/pulse/cyber-briefing-20251208-cybermaterial-6ugrc[28] React2Shell CVE-2025-55182: What it is and what to do – Dynatrace https://www.dynatrace.com/news/blog/cve-2025-55182-react2shell-critical-vulnerability-what-it-is-and-what-to-do/[29] Agfa-Gevaert Group Data Breach in 2025 – Breachsense https://www.breachsense.com/breaches/agfa-gevaert-group-data-breach/[30] VS: slachtoffers van ransomware betaalden 3,9 miljard euro losgeld https://www.security.nl/posting/916317/VS:+slachtoffers+van+ransomware+betaalden+3,9+miljard+euro+losgeld[31] Bleeping Computer News | Latest News on Bleeping Computer – Times of India https://timesofindia.indiatimes.com/topic/bleeping-computer/news[32] Detecting React2Shell: The maximum-severity RCE Vulnerability … https://www.sysdig.com/blog/detecting-react2shell[33] NCTV hekelt gebrek aan cybersecurity bij Rijk: ‘Zorgt voor vals … https://www.security.nl/posting/914777/Re:+NCTV+hekelt+gebrek+aan+cybersecurity+bij+Rijk:+’Zorgt+voor+vals+gevoel+van+veiligheid’[34] News in the Security category – Bleeping Computer https://www.bleepingcomputer.com/news/security/[35] React2Shell (CVE-2025-55182) https://react2shell.com[36] VK scherpt cybersecuritywetgeving aan voor vitale sectoren https://www.security.nl/posting/912559/VK+scherpt+cybersecuritywetgeving+aan+voor+vitale+sectoren[37] BleepingComputer | Cybersecurity, Technology News and … https://www.bleepingcomputer.com[38] React2Shell (CVE-2025-55182): Critical React Vulnerability | Wiz Blog https://www.wiz.io/blog/critical-vulnerability-in-react-cve-2025-55182[39] Agfa investigates alleged cybersecurity incident https://www.agfa.com/corporate/news-item/agfa-investigates-alleged-cybersecurity-incident/

DjimIT Nieuwsbrief

AI updates, praktijkcases en tool reviews — tweewekelijks, direct in uw inbox.