Can Information-Providing Artificial Intelligence Fight Disinformation? ChatGPT Example

Aytaç Burak Dereli; Erdem Taşdemir; Hilal Sevimli

Makale detayları

Can Information-Providing Artificial Intelligence Fight Disinformation? ChatGPT Example

Yazarlar:

Aytaç Burak Dereli | Erdem Taşdemir | Hilal Sevimli

Yayıncı:

Ondokuz Mayıs University

Ana Başlıklar:

İletişim Çalışmaları | İletişim Sistemleri | Yeni İletişim Teknolojileri

Dergi:

Middle Black Sea Journal of Communication Studies Vol.9 Issue 2, pp.106-133

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001331897

pISSN:

2587-1544

Anahtar kelimeler:

Özet:

 With the development of communication technologies, news can easily reach the masses without being monitored by any control center. Thus, the masses can quickly access unlimited content; they become vulnerable to disinformation that information density will bring. Disinformation, which spreads very quickly like information, can also cause significant problems in the public eye. However, artificial intelligence, which is one of the sources that enables the production and spread of disinformation, also plays an active role in detecting disinformation. This mission of artificial intelligence creates the need for how to use it in the most effective and correct way in order to prevent problems that may occur in digital media. ChatGPT, which shows a performance close to the human mind, is also an important medium that is frequently evaluated in the fight against disinformation. In this direction, the study aims to contribute to the literature by answering the question of whether artificial intelligence, which is an important source of disinformation production and dissemination, can detect news texts that are disinformation. The study was carried out by using content analysis, one of the qualitative research methods. Purposeful sampling was used for the purpose of the study; the extent to which news texts verified as fake news by the Directorate of Communications Disinformation Combat Center were detected by ChatGPT, an artificial intelligence chatbot, was categorized and analyzed. As a result of the study, it was determined that ChatGPT was indecisive in its responses to the detection of disinformation; it was rational in terms of providing new information on the subject and not providing a clear verification of disinformation, and it had a guiding attitude in terms of directing the user to various sources.

 With the development of communication technologies, news can easily reach the masses without being monitored by any control center. Thus, the masses can quickly access unlimited content; they become vulnerable to disinformation that information density will bring. Disinformation, which spreads very quickly like information, can also cause significant problems in the public eye. However, artificial intelligence, which is one of the sources that enables the production and spread of disinformation, also plays an active role in detecting disinformation. This mission of artificial intelligence creates the need for how to use it in the most effective and correct way in order to prevent problems that may occur in digital media. ChatGPT, which shows a performance close to the human mind, is also an important medium that is frequently evaluated in the fight against disinformation. In this direction, the study aims to contribute to the literature by answering the question of whether artificial intelligence, which is an important source of disinformation production and dissemination, can detect news texts that are disinformation. The study was carried out by using content analysis, one of the qualitative research methods. Purposeful sampling was used for the purpose of the study; the extent to which news texts verified as fake news by the Directorate of Communications Disinformation Combat Center were detected by ChatGPT, an artificial intelligence chatbot, was categorized and analyzed. As a result of the study, it was determined that ChatGPT was indecisive in its responses to the detection of disinformation; it was rational in terms of providing new information on the subject and not providing a clear verification of disinformation, and it had a guiding attitude in terms of directing the user to various sources.

 With the development of communication technologies, news can easily reach the masses without being monitored by any control center. Thus, the masses can quickly access unlimited content; they become vulnerable to disinformation that information density will bring. Disinformation, which spreads very quickly like information, can also cause significant problems in the public eye. However, artificial intelligence, which is one of the sources that enables the production and spread of disinformation, also plays an active role in detecting disinformation. This mission of artificial intelligence creates the need for how to use it in the most effective and correct way in order to prevent problems that may occur in digital media. ChatGPT, which shows a performance close to the human mind, is also an important medium that is frequently evaluated in the fight against disinformation. In this direction, the study aims to contribute to the literature by answering the question of whether artificial intelligence, which is an important source of disinformation production and dissemination, can detect news texts that are disinformation. The study was carried out by using content analysis, one of the qualitative research methods. Purposeful sampling was used for the purpose of the study; the extent to which news texts verified as fake news by the Directorate of Communications Disinformation Combat Center were detected by ChatGPT, an artificial intelligence chatbot, was categorized and analyzed. As a result of the study, it was determined that ChatGPT was indecisive in its responses to the detection of disinformation; it was rational in terms of providing new information on the subject and not providing a clear verification of disinformation, and it had a guiding attitude in terms of directing the user to various sources.

 İletişim teknolojilerinin gelişmesi ile haberler, herhangi bir kontrol merkezinde denetlenmeden kolayca kitlelere ulaşabilmektedir. Böylelikle kitleler sınırsız sayıda içeriğe hızlıca erişim sağlamakta; bilgi yoğunluğunun beraberinde getireceği dezenformasyona karşı savunmasız kalmaktadırlar. Enformasyon gibi oldukça hızlı yayılan dezenformasyon da kamu nezdinde önemli problemlere neden olabilmektedir. Bununla birlikte dezenformasyonun üretimi ve yayılmasını sağlayan kaynaklarından biri olan yapay zekâ, aynı zamanda dezenformasyonun tespit edilmesinde de etkin rol oynamaktadır. Yapay zekânın sahip olduğu bu misyon, dijital mecralarda yaşanabilecek problemlerin engellenebilmesi adına ondan en etkili ve doğru biçimde nasıl yararlanılması gerektiği ihtiyacını doğurmaktadır. İnsan zihnine yakın bir performans gösteren ChatGPT de dezenformasyonla mücadele konusunda sıklıkla değerlendirme altında olan önemli bir mecradır. Bu doğrultuda çalışma, dezenformasyon üretimi ve yayımının önemli kaynaklarından olan yapay zekâ, dezenformasyon niteliğindeki haber metinlerinin tespitini sağlayabilir mi, sorusuna cevap oluşturarak alanyazına katkı yapmayı amaçlamaktadır. Çalışma, nitel araştırma yöntemlerinden içerik analizi yapılarak gerçekleştirilmiştir. Çalışmanın amacı doğrultusunda amaçlı örneklem kullanılmış; İletişim Başkanlığı Dezenformasyonla Mücadele Merkezi tarafından sahte haber olarak doğrulanmış haber metinlerinin bir yapay zekâ sohbet robotu olan ChatGPT tarafından ne derece tespit edilebildiği kategorize edilerek analiz edilmiştir. Çalışmanın sonucunda ChatGPTnin dezenformasyonun tespitine yönelik yanıtlarda kararsız kaldığı; konuya ilişkin yeni bilgiler sunması ve dezenformasyona ilişkin net bir doğrulama yapmaması açısından rasyonel, kullanıcıyı çeşitli kaynaklara sevk etmesi açışından yönlendirici bir tutuma sahip olduğu saptanmıştır.

Makale detayları

Can Information-Providing Artificial Intelligence Fight Disinformation? ChatGPT Example

Yazarlar:

Aytaç Burak Dereli | Erdem Taşdemir | Hilal Sevimli

Yayıncı:

Ondokuz Mayıs University

Ana Başlıklar:

İletişim Çalışmaları | İletişim Sistemleri | Yeni İletişim Teknolojileri

Dergi:

Middle Black Sea Journal of Communication Studies Vol.9 Issue 2, pp.106-133

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001331897

pISSN:

2587-1544

Disinformation

News

Artificial Intelligence

ChatGPTDezenformasyon

Haber

Yapay Zekâ

ChatGPT

Sayı

Makale detayları

Yazarlar: Aytaç Burak Dereli | Erdem Taşdemir | Hilal Sevimli

Yayıncı: Ondokuz Mayıs University

Ana Başlıklar: İletişim Çalışmaları | İletişim Sistemleri | Yeni İletişim Teknolojileri

Dergi: Middle Black Sea Journal of Communication Studies Vol.9 Issue 2, pp.106-133

Belge Tipi: Makale

Yayın Tarihi: 12/2024

Belge Dili: Türkçe

Document ID (Org No.): 2001331897

pISSN: 2587-1544

Sayı

Yazarlar:

Aytaç Burak Dereli | Erdem Taşdemir | Hilal Sevimli

Yayıncı:

Ondokuz Mayıs University

Ana Başlıklar:

İletişim Çalışmaları | İletişim Sistemleri | Yeni İletişim Teknolojileri

Dergi:

Middle Black Sea Journal of Communication Studies Vol.9 Issue 2, pp.106-133

Belge Tipi:

Makale

Yayın Tarihi:

12/2024

Belge Dili:

Türkçe

Document ID (Org No.):

2001331897

pISSN:

2587-1544