AI Sec Watch

Large language models (LLMs, AI systems trained on vast amounts of text to predict and generate human-like language) like ChatGPT can help with tasks like drafting and summarizing, but they may produce incorrect information or outdated answers since they rely on patterns in their training data rather than real-time information. To use these tools safely, you should verify important facts with trusted sources, check for bias in outputs, seek advice from qualified professionals for legal or medical decisions, and be transparent about your AI use in work or school settings.

This content is a reference guide showing how ChatGPT can assist managers across ten different job areas, from strategy planning to crisis management. For each area (like hiring, performance reviews, or decision-making), it lists example scenarios and the types of documents or frameworks ChatGPT can help produce. This is a tool overview, not a discussion of AI risks or technical issues.

This document describes how marketing teams can use ChatGPT, an AI language model, to speed up their work across campaigns, content creation, and performance analysis. ChatGPT helps teams move from initial ideas through drafting and launch by organizing scattered inputs into clear messaging, generating content variations, and summarizing performance data. The tool is most effective when treated as a thinking partner for iterative refinement rather than a one-time solution, with human judgment applied for final decisions.

This document outlines how ChatGPT can assist sales teams by generating helpful outputs for various stages of the sales process, from initial prospecting and research through deal closure. It covers practical applications like creating account briefs, discovery guides, meeting agendas, email sequences, proposals, and objection-handling talk tracks across eight common sales scenarios.

Prompt engineering is the process of designing and refining your input to help ChatGPT give better answers. The document explains that clear prompts work best when you specify what you need, provide relevant context, describe the desired output format, and break complex tasks into smaller steps. There is no single perfect way to write a prompt, so experimentation and iteration help you discover how to use AI most effectively.

AI is software that recognizes patterns and learns from data to produce useful outputs, with large language models (LLMs, systems trained on large amounts of text to generate and transform language) being a common type you interact with through tools like ChatGPT. Models go through two training stages: pre-training, where they learn general patterns from massive text datasets, and post-training, where they're refined to follow instructions reliably, communicate clearly, and handle sensitive topics carefully through safety checks. Different models are optimized for different tradeoffs, such as reasoning models designed for complex problem-solving versus non-reasoning models built for fast, straightforward tasks.

Custom GPTs are tailored versions of ChatGPT built for specific, repeatable tasks, where you define how the GPT behaves through instructions and can add knowledge (uploaded documents) and tools like web search or data analysis. They work best when you find yourself reusing the same prompts or instructions across multiple tasks, reducing repetition and keeping context consistent. You create a custom GPT by opening the GPT builder in ChatGPT, naming it, writing clear instructions for how it should behave, and optionally uploading files or enabling features like image generation or code analysis.

OpenAI has released features that let you customize how ChatGPT behaves by using custom instructions (settings that tell ChatGPT about your role and preferred communication style) and memory (which stores information you want ChatGPT to remember across conversations). These personalization tools help ChatGPT work more like a reliable teammate by building context over time, so you don't have to repeat the same information every time you chat.

OpenAI discovered that Axios, a third-party developer library (a pre-written code package used to build software), was compromised in a software supply chain attack (where attackers infiltrate widely-used tools to affect many companies at once) on March 31, 2026, and their macOS app-signing process briefly used a malicious version. OpenAI found no evidence that user data or systems were compromised, but is revoking and updating their security certificates (digital credentials that verify software is authentic) and requiring all macOS users to update their OpenAI apps to prevent the risk of fake apps appearing legitimate. As of May 8, 2026, older versions of ChatGPT Desktop (before 1.2026.051), Codex App (before 26.406.40811), Codex CLI (before 0.119.0), and Atlas (before 1.2026.84.2) will no longer receive updates and may stop working.