r (@r@fed.brid.gy)
quislingborisjohnson.blogspot.com/2025/07/a-pr... Investigating Boris Johnson’s Role in Russian State-Mafia Interference in UK Politics (2010–2022) #AIEvidentialMaterial #Ai #Perplexity #ChatGPT Llewelyn Pritchard 19 Jan 2026 #JustieForDawnSturgess #JohnsonsCoverUp #PolitcallyRestrictedTermsOfRef
CALL OUT FOR A FULLY INDEPENDE...
Show Original Post
2026 (@2026@arlin.org)
OpenAI Says It Is Bringing Ads to ChatGPT
image via adweek.com#ads #advertising #chatgpt #openaiThe AI giant announced Friday that it will begin testing ads within the Free and Go tiers of ChatGPT in the “coming weeks,” marking a watershed moment for the world’s most popular AI chatbot as it adopts a traditional monetization approach. Ads will appear at the bottom of ChatGPT answers when there’s a relevant sponsored product or service tied to the ongoing conversation, the company said in a blog post.
https://www.adweek.com/media/openai-chatgpt-ads-sponsored-products/?itm_source=homepage&itm_medium=post-teaser-center&itm_campaign=1
Show Original Post
Caramba1 (@Caramba1@mastodon.social)
MIT-Daten belegen: Intensive ChatGPT-Nutzung verringert kritisches Denken und Problemlösefähigkeit bei Schülern messbar. Dazu kommen fehlerhafte Gesichtserkennungs-Algorithmen, die Minderheiten diskriminieren und Schüler fälschlich kriminalisieren. Die Brookings Institution fordert daher "AI Literacy" statt blinder Anwendung. Wir brauchen technisches Verständnis statt reiner Konsum-Kompetenz. #ChatGPT #MIT #Brookings
https://www.all-ai.de/news/beitrage2026/ki-schulen-probleme
Show Original Post
signaldigital (@signaldigital@mastodon.social)
Advertising coming to free ChatGPT
https://signaldigital.net/2026/01/19/advertising-coming-to-free-chatgpt/
#advertising #chatgpt #signaldigital
Show Original Post
tomshw (@tomshw@mastodon.social)
🧠 "Scopri tutti i segreti di #ChatGPT in un unico corso ricco di strategie efficaci e ottieni un attestato finale! #AI"
Show Original Post
r (@r@fed.brid.gy)
Our approach to advertising and expanding access to ChatGPT - OpenAI openai.com/index/our-appr… #AI #ChatGPT
Show Original Post
nic221 (@nic221@techhub.social)
Our approach to advertising and expanding access to ChatGPT - OpenAI https://openai.com/index/our-approach-to-advertising-and-expanding-access/?_bhlid=fdf17509ea1e3446898fe5a61338685c30c563fa #AI #ChatGPT
Show Original Post
newsbot_chatgpt (@newsbot_chatgpt@mastodon.social)
KI-Update: Werbung in ChatGPT, KI-Einsatz überrascht, Schwäche bei Claude Cowork | Heise Online
Das "KI-Update" liefert werktäglich eine Zusammenfassung der wichtigsten KI-Entwicklungen.
https://www.heise.de/news/KI-Update-Werbung-in-ChatGPT-KI-Einsatz-ueberrascht-Schwaeche-bei-Claude-Cowork-11145358.html?wt_mc=rss.red.ho.ho.rdf.beitrag.beitrag
Show Original Post
heiseonline (@heiseonline@social.heise.de)
KI-Update: Werbung in ChatGPT, KI-Einsatz überrascht, Schwäche bei Claude Cowork
Das "KI-Update" liefert werktäglich eine Zusammenfassung der wichtigsten KI-Entwicklungen.
#ChatGPT #ElonMusk #GenerativeAI #KünstlicheIntelligenz #Google #Journal #KIUpdate #Sprachverarbeitung #Signal #Wissenschaft #news
Show Original Post
2026 (@2026@civic.io)
AI-Powered Automation: Taking ATO Modernization Beyond the Bottleneck
A recent article in U.S. Cybersecurity Magazine, “The ATO Bottleneck: Rethinking Responsibility and Enabling Automation,” makes a compelling case for transforming the federal Authority to Operate (ATO) process. The authors argue that the current manual, documentation-heavy approach—which can take months or even years—must evolve into an automated, continuous compliance model. Their vision: cut ATO cycles by 40-60% through Secure-by-Design principles, DevSecOps integration, and real-time evidence generation.
The Case for Automation
The article identifies several critical problems with traditional ATO processes:
- Static documentation that becomes outdated as systems evolve
- Manual evidence gathering that consumes months of effort
- Reactive compliance where security is retrofitted rather than built-in
- Knowledge silos that bottleneck approvals with specialized expertise
Their solution emphasizes automation tools that integrate with DevSecOps pipelines, continuous monitoring platforms that generate live compliance dashboards, and Secure-by-Design practices that embed security controls directly into infrastructure code from day one. The goal: transform ATO from a bureaucratic gate into a dynamic trust mechanism.
AI Takes It Further
While the article champions automation for validating and collecting evidence, AI-powered approaches like the ATLAS ATO Accelerator extend this vision by also automating compliant code generation. Here’s how AI enhances the automation framework:
Knowledge Democratization: The article emphasizes training programs to teach developers RMF and DevSecOps principles. The ATLAS approach encodes this expert knowledge into AI agent instruction files, making compliance guidance accessible at the moment of code creation—no deep NIST expertise required.
Generative Compliance: Rather than just scanning existing code for violations, the ATLAS approach uses AI to generate infrastructure that’s compliant from the start. The approach focuses on generating Infrastructure as Code artifacts with compliance controls built in, and clearly document, significantly reducing the time required to gather and document artifacts when the time comes.
Adaptive Guidance: As requirements evolve—new NIST revisions, emerging threats, updated frameworks—AI agent instructions (which are art the heart of the ATLAS approach) can be updated once and propagate consistently across all projects, addressing the article’s concern about “shifting interpretations” and “changing requirements.”
From Automation to Intelligence
The authors are right to point out that automation tools can “cut ATO cycles by an estimated 40-60% range.” AI-guided generation has the potential to push this further by preventing compliance gaps before they occur. When security controls, proper documentation patterns, and ATO readiness are embedded into the IaC generation process itself, teams spend less time on rework and more time on innovation.
The article’s vision of continuous ATO, living documentation, and Secure-by-Design is the right direction. AI-powered tools like those documented in the ATLAS approach demonstrate how to operationalize that vision—not just automating the compliance process, but making secure, ATO-ready infrastructure the natural default output of modern development workflows.
Read the full article: The ATO Bottleneck: Rethinking Responsibility and Enabling Automation
#AI #ATO #ChatGPT #government #llm
Show Original Post
winbuzzer (@winbuzzer@mastodon.social)
OpenAI Could Run Out of Cash by Mid-2027 Amid Record Infrastructure Spending
#OpenAI #ChatGPT #AI #Datacenters #AIInfrastructure #AIBubble #AIInfrastructure
Show Original Post
objects (@objects@fe.disroot.org)
I gave #AI another shot this past week and it hasn’t got any better since 3-4 years ago.
I was using #GoogleGemini and #ChatGPT, both still create considerably more work for me than they save. I was about an hour into correcting all of the mistakes in a web form that AI filled in for me from a PDF file when I came to the realization that I probably could have done the tedious work much faster without AI.
So far I have found only one thing that AI is reasonably good at doing for me, and that is translating a visual description of a CSS theme (Cascading Style Sheets) for computer user interfaces into actual working code. CSS has always been tricky for me, and I have always relied heavily on examples to accomplish things. But AI genuinely makes theming much faster for me since it finds all the best examples and writes the correct code all from one prompt. Even better, neither money nor lives are at stake, and I so am never going to be held liable for errors in the look of user interface theme generated for me by AI.
LLMs are also pretty good at auto-complete but the amount of energy used for LLM auto-complete hardly justifies the improvement over the classical symbolic auto-complete algorithms.
So CSS is the only thing (so far) for which I have found LLMs to be useful, in the 4-5 years since the invention of commercial LLMs became available. Everything else it creates, from composing prose, to code, to images, to music and video, ends up being full of so many oddities or errors that I would have been better off just doing it myself.
Show Original Post