A perícia forense na era do Deepfake e da IA | Starti Cast #08

A perícia forense na era do Deepfake e da IA | Starti Cast #08

🎙 Starti 👥 33K 📅 June 26, 2026 ⏱ 57 min 👁 142 📄 expert opinion 🧭 2026-08-16
Available in: English (current) Français

Keywords

deepfakeforensiccybersecurityAIdigital evidence

Summary

In this episode of Starti Cast, host Mirem interviews Ricardo Capozzi, a forensic computing expert and technical assistant specializing in cybercrimes, image analysis, and audio transcription. The conversation explores how AI and deepfakes have transformed digital crime and forensic investigation. Capozzi explains his background in computing and law, highlighting the importance of combining technical and legal knowledge. He discusses the challenges of analyzing manipulated audio and video, emphasizing the need for proper chain of custody and metadata analysis. He illustrates how AI can be used both to create convincing fakes and to detect them, using techniques like analyzing idiosyncrasies and pixel-level details. The episode also covers practical advice for businesses, such as protecting against identity fraud and understanding the legal implications of digital evidence. Capozzi stresses the lack of a security culture and the importance of behavioral changes over relying solely on technology. The discussion includes real-world examples, such as an inconclusive forensic analysis due to poor image quality, and explains the difference between criminal and civil forensic investigations. The episode concludes with advice on how companies can safeguard themselves against AI-driven fraud.

183 words

Critical Evaluation

Value of the Information & Strength of the Argument

The value of the information is high for a general audience interested in cybersecurity and digital forensics. The expert provides practical insights into how deepfakes are created and detected, and the importance of legal procedures. The argumentation is solid, based on the expert’s professional experience and legal training. He clearly explains complex concepts like chain of custody and the limitations of AI in forensic analysis. However, some claims are anecdotal and lack empirical data or citations. The discussion is persuasive and well-structured, but could benefit from more concrete examples and references.

Scientific Rigor, Source Quality, Title Accuracy

The scientific rigor is moderate. The expert demonstrates deep knowledge of the subject, but the content is primarily opinion-based rather than research-driven. No formal sources are cited within the video, and the description provides only links to the company’s blog and social media, which are not academic references. The title accurately reflects the content, which is a focused discussion on forensic expertise in the era of deepfakes and AI. The lack of citations and reliance on personal experience limit the scientific robustness, but the practical insights are valuable.

194 words

Title / Content Match

The title accurately reflects the content, which focuses on forensic expertise in the era of deepfakes and AI.

Quality & Reliability

7/10

The content is an expert interview with a forensic computing expert, providing practical insights into digital forensics and deepfake analysis. The information is based on professional experience and legal knowledge, but lacks formal citations or peer-reviewed references. The discussion is coherent and grounded in real-world cases, though some claims are anecdotal.

Key Moments

Cited Sources

Concurring Sources

Contribution & Novelties

The video provides a practical, expert perspective on the intersection of AI, deepfakes, and forensic investigation. It demystifies the process of digital forensics and highlights the challenges of proving authenticity in legal contexts. The discussion on the importance of chain of custody and the limitations of AI in providing conclusive evidence is particularly valuable for businesses and legal professionals.

Pour aller plus loin :

128 words

Radar Profile

The radar profile shows a balanced performance across all dimensions, with slightly higher scores in information quality and reliability, reflecting the expert's credibility. The lower score in technical level indicates that the content is accessible to a general audience, while still providing valuable insights.

Reliability 7/10

💬 No comments were provided for analysis.