Anderson’un AÃ§ÄąsÄą

Yapay Zeka ‘Pembe Çamur’ Haberlerini TanÄąyabilir

mm
Unite.AI sitesini Google'daki tercih ettiğiniz kaynaklara ekleyin
AI-generated illustration featuring two 'perfect' and typical US-style newsreaders in a news anchor room – an older male and younger blonde female both Caucasian, with robots and technicians in the background. Z-Image, Firefly 3, et al.

Agenda-driven opinion mills, designed more to sway public opinion than serve the public, might be harder to spot if AI is used to make them sound more original and rational. So the race is on to stay ahead in the ‘pink slime detection’ game.

 

Son yirmi yÄąlda geleneksel yerel medya kuruluşlarÄąnÄąn fonlarÄąnÄąn kesilmesi, hem değişen medya trendleri hem de – son zamanlarda – ABD hÞkÞmeti politikasÄą nedeniyle, bÃķlgesel raporlamada bir boşluk oluşmuş ve bu boşluk partizan organizasyonlar tarafÄąndan yapay zeka kullanarak gÞndemlerini ilerletmek için aceleyle doldurulmuştur.

‘Partizan’ terimini bağlam içinde kullanmak gerekirse (çÞnkÞ hiçbir haber kuruluşu siyasi eğilimlerden tamamen uzak değildir), petrol şirketlerinin yerel haber sitelerini uzak yerlerden yÃķnettiğini, yerel kaynaklara sahip olmadığınÄą, ancak şirketin kamu imajÄąnÄą savunma gÃķrevi olduğunu, seçimlerden Ãķnce siyasi olarak motive edilmiş haber sitelerinin gelir akÄąÅŸÄąndan yoksun olduğunu ve Cumhuriyetçi haber sitelerinin hiçbir yerden ortaya Ã§ÄąktığınÄą konuşuyoruz.

2024 yÄąlÄąnda, yapay zeka destekli pembe çamur haberlerinin gerçek haber sitelerini nihayet aşmÄąÅŸ olabileceği tahmin edildi; o zaman, bir Avustralya anketi, tÞketicilerin %41’inin pembe çamur kaynaklarÄąnÄą “gerçek” kaynaklardan daha çok tercih ettiğini buldu.

Bu tÞr gizli seçim propagandasÄą, demokrasiye (siyasi olarak motive edilmiş siteler için) ve haberlerin makul standartlarÄąnda adil olunmasÄąna gÞven için bir varoluşsal tehdit olarak gÃķrÞlebilir.

Pembe çamur yayÄąncÄąlarÄą ve yayÄąncÄąlarÄąnÄąn karakteristik Ã§ÄąktÄąsÄąnÄą geleneksel medya kuruluşlarÄąndan ayÄąrt etmek için yÃķntemler geliştirilmesi bÞyÞk yardÄąm olurdu.

As it stands, the tropes and templates of real news organizations are very easy to mimic, and AI makes scalable publishing a current and affordable reality, using many of the same tricks being adopted by budget-stricken ‘old media’ publishers and broadcasters.

Sinyal ve GÞrÞltÞ

A new study from the US addresses this issue, by investigating the growing use of Large Language Models to make pink slime websites sound less generic and easy-to-spot, and by creating a learning framework designed to keep up with evolving changes in pink slime (PS) output.

Titled Pembe Çamur Gazeteciliğini AÃ§ÄąÄŸa Ã‡Äąkarmak: Dilsel İmzalar ve LLM Oluşturulan Tehditlere KarÅŸÄą DayanÄąklÄą Tespit, the new work comes from five researchers at the University of Texas.

The new work investigates how mass-produced PS local news articles differ from legitimate reporting, focusing on their reliance on short, repetitive structures and templated phrasing with minimal variation; and the authors note that PS articles tend to reuse identical templates designed to manipulate public opinion, with appeals to emotion uppermost in the content:

From the new paper - multiple outlets publish near-identical articles with only location details changed, revealing a copy-paste strategy used to mass-produce content that mimics legitimate local news. Source - https://arxiv.org/pdf/2512.05331

Yeni çalÄąÅŸmadan – birden fazla yayÄąn aynÄą içerikleri sadece konum ayrÄąntÄąlarÄąnÄą değiştirerek yayÄąnlÄąyor, bu da meşru yerel haberleri taklit eden içerikleri kitleler halinde Þretmek için kullanÄąlan bir kopyala-yapÄąÅŸtÄąr stratejisi ortaya koyuyor. Kaynak

Traditional detection models trained on these traits perform well against such content, but fail when the articles are rewritten using AI chatbots to appear more natural or sophisticated.

The authors’ own tests indicate that even minor stylistic changes introduced by large language models can reduce detection accuracy by up to 40%. To mitigate this, they propose a continual learning framework that incrementally retrains detection models on both original and AI-rewritten articles, to adapt to shifting linguistic patterns.

YÃķntem

To establish data for the project, the authors used the Pembe Çamur Veri Seti, featuring 7.9 million articles covering 1,093 outlets during 2021-2023, from which they obtained 9,472 pink slime articles after filtering. They also used the LIAR dataset, which contains annotated fake news, as well as the NELA-GT-2021 collection, which contains solely US articles*.

To prepare their training and testing sets, the authors first used the T-distributed Stochastic Neighbor Embedding (t-SNE) algorithm to reduce the article embeddings to two dimensions. They then applied the data clustering algorithm Density-Based-Spatial-Clustering-of-Applications-with-Noise (DBSCAN) to isolate clusters of similar pink-slime articles.

Each cluster was treated as a group of related stories, many of which still followed the same template, despite a concerted effort to address duplicates.

To prevent similar articles from appearing in both training and test sets, entire clusters were randomly selected, with 80% used for training and 20% for testing. Because the legitimate news articles did not form clear clusters, a random split was applied instead.

This process was repeated three times, to ensure consistency, and to reduce sampling bias.

Pembe Çamurun Özellikleri

Commenting on the distinguishing traits of PS vs. regular news, the researchers assert that PS-style local news articles are significantly shorter and simpler than legitimate reporting, averaging fewer than dokuz cÞmle per article.

A higher proportion of simple sentences and a heavier reliance on adjectives are further hallmarks of pink slime, according to the paper, and indicate a penchant for repetitive, emotionally charged language.

Lexical richness was measured using the Root-Type-Token Ratio (RTTR), and found to be notably lower in the PS articles, which also exhibited far fewer unique noun phrases.

These patterns denote a limited vocabulary and formulaic style, in contrast to legitimate local news, which is characterized by complex part-of-speech patterns built around auxiliary verbs, pronouns, and conjunctions. Instead, the fake articles favor basic noun-preposition structures, with frequent use of punctuation-based trigrams, suggesting a less formal, more fragmented writing style.

Testler

To examine the associations between different types of news articles, based on linguistic and structural features, embeddings were generated using the 435-million parameter stella_en_400M_v5 model, and reduced with Principal Component Analysis (PCA), and t-SNE for visualization.

When projected into two dimensions, the fake local news articles formed small, dense clusters, each corresponding to narrowly focused topics such as crime statistics, stock updates, or charitable donations:

Clustering patterns from a t-SNE projection reveal that pink slime articles form tight, repetitive groupings, while legitimate news displays broader, more varied distributions aligned with topic and style diversity.

t-SNE projeksiyonundan elde edilen kÞmeleme kalÄąplarÄą, pembe çamur makalelerinin sÄąkÄą, tekrarlÄą gruplar oluşturduğunu, meşru haberlerin ise daha geniş, daha çeşitli dağılÄąmlar sergilediğini gÃķsterir.

As we can see, to some extent, in the visualization above, this pattern suggests a rigid, template-driven format, with minimal variation between articles.

Interestingly, articles labeled as ‘fake news’ diverged from the fake local content, showing a distribution more in line with real news, indicating that mass-produced local fakes may not be simply less truthful, but may also be mechanically distinct in form and composition.

By contrast, ‘legitimate’ local news forms fewer and more widely spaced clusters, consistent with more diverse language and subject matter, while national news articles show even greater dispersion, reflecting broader topical range and looser stylistic consistency.

Feature comparisons between legitimate local news and pink slime content, indicating that PS articles are shorter, use simpler sentence structures, contain more adjectives, exhibit lower lexical richness, favor basic part-of-speech trigrams, and contain fewer unique noun phrases.

Meşru yerel haberler ve pembe çamur içeriği arasÄąndaki Ãķzellikler karÅŸÄąlaştÄąrmasÄą, PS makalelerinin daha kÄąsa olduğunu, daha basit cÞmle yapÄąlarÄąnÄą kullandığınÄą, daha fazla sÄąfat içerdiğini, daha dÞşÞk sÃķzcÞk zenginliğine sahip olduğunu, temel sÃķzdizimi ÞçlÞlerini tercih ettiğini ve daha az benzersiz isim terkibine sahip olduğunu gÃķsterir.

Tespit

The researchers evaluated two main approaches to detect pink slime content: classification, based on handcrafted linguistic features; and transformer-based fine-tuning.

For the handcrafted approach, structural rather than semantic characteristics were emphasized, using sentence count; lexical richness; syntactic depth; part-of-speech co-occurrence probabilities; dependency tag co-occurrence probabilities; readability; and part-of-speech counts.

Three models were tested on this feature set: XGBoost; Random Forest; and Support Vector Machine (SVM) – with Random Forest showing slightly stronger results overall.

Both XGBoost and Random Forest assigned high predictive importance to features such as sentence count and the number of unique noun phrases. Readability and lexical richness measures also influenced classification strongly, though the models weighted these differently, with XGBoost favoring Flesch and RTTR, while Random Forest leaned on CTTR:

Feature importance scores based on SHAP (SHapley Additive exPlanations) highlight how each input feature influences the model’s output across samples. In this case, SHAP values reveal that both XGBoost and Random Forest relied most heavily on sentence count and unique noun phrases to distinguish pink slime from real news, while assigning varying weight to measures of lexical richness and readability.

SHAP (SHapley Additive exPlanations) temelli Ãķzellik Ãķnem puanlarÄą, her bir girdi Ãķzelliğinin Ãķrnekler genelinde model Ã§ÄąktÄąsÄąnÄą nasÄąl etkilediğini vurgular. Bu durumda, SHAP değerleri, XGBoost ve Random Forest’un her ikisinin de pembe çamuru gerçek haberden ayÄąrmak için cÞmle sayÄąsÄąna ve benzersiz isim terkibine en çok gÞvendiğini gÃķsterirken, sÃķzcÞk zenginliği ve okunabilirlik ÃķlçÞmlerine farklÄą ağırlÄąk verdiğini gÃķsterir.

As seen in the earlier feature comparisons (above), pink slime articles favor sensationalism over detail, with a lower lexical richness, and with fewer distinct noun phrases – reinforcing the conclusion that this kind of content is highly templated and repetitive.

Part-of-speech trigram patterns confirm that legitimate local news tends to use more structurally complex forms, involving auxiliaries, pronouns, and conjunctions, while pink slime favors fragmented or minimal syntax. These patterns suggest a reliable foundation for distinguishing synthetic local content from real journalism.

A second round of testing involved fine-tuning transformer models on full article text, to capture both semantic content and syntactic structure.

BERT, XLNet, and Flan-T5 were tested, with BERT achieving the highest F1-score at 89.31% – although the paper notes that performance differences were not statistically significant. Unlike the handcrafted classifiers, these models update all weights during training, allowing them to learn task-specific representations directly from the data.

An additional embedding-based method used a fully-connected downstream classifier performed better than the handcrafted models, but did not match the accuracy of full fine-tuning.

These approaches benefited from broader linguistic context – advantageous, given that pink slime articles often recycle content across outlets. Handcrafted features, by contrast, rely on surface-level syntax alone, limiting their ability to generalize across sources.

LLM’lerle Pembe Çamurun GÞçlendirilmesi

To test whether pink slime articles could be made harder to detect, the researchers rewrote them using large language models. Prompts were designed to improve the writing style by expanding content, adding sentence complexity, reducing overuse of adjectives, enforcing a neutral tone, using more varied vocabulary, and increasing the number of unique noun phrases, with the ambit of hiding the surface-level traits that most detection systems rely on.

The models tested were Gemma‑2B; Mistral‑7B; Llama‑1B; Llama‑3B; Llama‑8B; Qwen‑7B; GPT‑4o mini; and Claude‑3.5‑Haiku.

Detection performance (F1) of a fine-tuned BERT classifier when pink slime articles are rewritten by different LLMs. Lower scores indicate more successful obfuscation.

FarklÄą LLM’ler tarafÄąndan yeniden yazÄąlmÄąÅŸ pembe çamur makalelerinde ince ayarlanmÄąÅŸ bir BERT sÄąnÄąflandÄąrÄącÄąnÄąn tespit performansÄą (F1). Daha dÞşÞk puanlar daha başarÄąlÄą gizleme anlamÄąna gelir.

Of these results, the authors comment:

‘[AÃ§Äąk kaynaklÄą modeller arasÄąnda] Qwen, tespitten kaÃ§Äąnmada en etkili olanÄąydÄą: daha Ãķnce ince ayarlanmÄąÅŸ bir BERT sÄąnÄąflandÄąrÄącÄąya (başlangÄąÃ§ F1 puanÄą: %89) karÅŸÄą test edildiğinde, tespit performansÄąnda %28,34’lÞk bir dÞşÞşe neden oldu.’

‘KapalÄą kaynaklÄą modeller çok daha gÞçlÞydÞ, GPT-4o-mini ve Claude-3.5-Haiku, F1 puanÄąnÄą ortalama %40 oranÄąnda dÞşÞrdÞ, bu da yÞksek kaliteli LLM oluşturulan gizlemenin oluşturduğu zorluğu vurguladÄą.’

These results, the authors contend, show how easily LLMs can disguise pink slime content, making it much harder for current tools to catch**.

Sonuç

GÃķrÞş Bu araştÄąrma hattÄą bazÄą ilginç ikilemler içerir, bunlardan en az biri, birçok insanÄąn (en az bir anket tarafÄąndan bahsedildiği gibi) desteklediği pembe çamur içeriği, bu da pejoratif bağlamÄą sorgulamaya yol açar. İnsanlarÄąn “Soylent Green insanlarÄąn eti” olduğunu bildikleri, ancak yine de yedikleri gibi gÃķrÞnÞyor; ya da liberal bir duyarlÄąlÄąktan bakÄąldığında bÃķyle gÃķrÞnÞyor.

Bu kamu ilgisizliği algoritmik haberlere karÅŸÄą derinleşiyor gibi gÃķrÞnÞyor.

Beni bu makaleyi okurken etkileyen bir diğer şey, pembe çamur Ã§ÄąkÄąÅŸÄąnÄąn daha basit Þslubu ve indirgeyiciliğinin bir eksiklik olarak gÃķrÞlmesi, ancak bu minimalizm, duygusallÄąk ve sÄąnÄąrlÄą sÃķzcÞk dağarcığınÄąn aslÄąnda kasÄątlÄą olduğudur.

Eğer pembe çamurun arkasÄąndaki çeşitli Ã§Äąkar gruplarÄą daha entelektÞel veya liberal bir kitleye ulaşmaya çalÄąÅŸÄąyorsa (bu, gÞçlerine gÃķre olmayabilir), mevcut platformlarda değil, hedef kitlelerine daha yakÄąn bir yerde kamp kurmalarÄą daha muhtemeldir.

 

* Maalesef, makaledeki bazÄą biçimlendirme sorunlarÄą nedeniyle, yerel haber makalelerinin ek kaynağı net bir atÄąfta bulunmuyor. LÞtfen kaynak makaleye başvurun ve hangi ‘Horne’ referansÄąnÄąn uygulanacağınÄą tahmin edin.

** Burada, sonuç bÃķlÞmÞnÞn sonundaki ikincil, tamamlayÄącÄą deneylerin ayrÄąntÄąlarÄą için okuyucuyu kaynak makaleye yÃķnlendiriyoruz.

İlk olarak 12 AralÄąk 2025 Cuma gÞnÞ yayÄąnlandÄą.

Makine Ãķğrenimi Þzerine yazar, insan gÃķrÞntÞ sentezi alanÄąnda uzman. Metaphysic.ai'de araştÄąrma içeriği başkanÄą, DNEG'nin Brahma.ai'ye dÃķnÞşÞmÞne kadar.
Portfolio sitesi: martinanderson.ai
İletişim: [email protected]