ChatGPT Can Now See Everything You Do On Your Screen

ChatGPT Can Now See Everything You Do On Your Screen

🎙 Igor Pogany 👥 480K 📅 August 14, 2026 ⏱ 15 min 👁 22K 📄 news review 🧭 2026-09-08
Available in: English (current) Français

Keywords

ChatGPTcomputer historyscreen recordingprivacyAI agents

Summary

The video presents OpenAI’s new ‘Computer History’ feature for the ChatGPT desktop app, which allows the AI to track and summarize user activity across applications and websites. The creator demonstrates the feature by performing a series of tasks (creating a document, using ChatGPT, sending an email) and then asking ChatGPT to summarize what was done, create a reusable skill, and suggest automatable tasks. The video highlights the feature’s potential to provide AI with rich contextual data, enabling more personalized assistance and automation. However, it also raises significant privacy and security concerns, noting the feature’s limited availability (Mac only, certain plans, not in Europe) and comparing it to the backlash against Microsoft’s Recall. The creator expresses excitement about the feature’s power but also advises caution and careful scoping of app access. The video concludes with a brief mention of other AI news, including a new GPT-5.6 model and Linux support for the ChatGPT app.

153 words

Critical Evaluation

Value of the Information & Strength of the Argument

The video provides a valuable, hands-on demonstration of a new AI feature, showing its practical applications and potential benefits. The argumentation is clear and persuasive, highlighting how the feature could simplify AI usage by providing context and enabling automation. The creator effectively uses a real-world example to illustrate the feature’s capabilities, making the information accessible and engaging. However, the argumentation is largely one-sided, focusing on the positive aspects while acknowledging but not deeply exploring the privacy risks. The video’s value lies in its timely and practical insight into a cutting-edge AI development.

Scientific Rigor, Source Quality, Title Accuracy

The video references the official ChatGPT documentation for the computer history feature, which adds credibility. However, it does not cite any independent sources or studies to support claims about the feature’s effectiveness or safety. The title accurately reflects the content, which is a review and demonstration of the feature. The video is a personal opinion and demonstration rather than a rigorous scientific analysis, but it provides useful information for users considering the feature.

180 words

Title / Content Match

The title accurately reflects the content, which focuses on ChatGPT's new computer history feature and its implications.

Quality & Reliability

7/10

The video provides a hands-on demonstration of a new feature, with clear explanations and references to official documentation. However, it lacks independent verification and relies heavily on the creator's personal experience and opinions.

Key Moments

Cited Sources

Concurring Sources

Dissenting Sources

  • Microsoft Recall backlash — The video mentions the backlash against a similar feature, suggesting potential public concern.

External References

Contribution & Novelties

The video provides a practical, first-hand look at a new AI feature that could significantly change how users interact with AI assistants. It highlights the potential for AI to gain deep context from user activity, enabling more proactive and personalized assistance. The video also raises important questions about privacy and the balance between convenience and data security.

Pour aller plus loin :

  • Microsoft Recall — A similar feature that faced backlash, providing context for comparison.
  • AI agent — Background on AI agents and their capabilities.
  • Data privacy — General concept of data privacy relevant to the discussion.

97 words

Radar Profile

The radar profile shows high scores in information quantity and quality, reflecting the video's detailed demonstration and clear explanation. The technical level is moderate, suitable for a general audience. The overall reliability is good, but limited by the lack of independent verification.

Reliability 7/10

💬 Sur les 0 commentaires analysés, aucune tendance n'est disponible.