AI/ML in the News - Highlights 15 Aug 2025

Posted on Aug 15, 2025

The articles cover major developments in AI models like OpenAI’s GPT-5 and applications across sectors from healthcare to content moderation, while raising concerns about AI’s societal impacts. Key issues for universities include AI’s effects on education and research practices, preparing students for an AI-driven workforce, and the ethical considerations around AI development and deployment.

There are conflicting views on how close we are to achieving artificial general intelligence (AGI), with some experts believing it is within reach in the next 5 years (Milmo and Kerr 2025) while others argue we are still far from true AGI (Orland 2025). The articles also present differing perspectives on whether AI will primarily augment human capabilities or replace human workers in many fields.

A. Developments in AI/ML models

  • OpenAI launched GPT-5, claiming significant improvements in reasoning capabilities but acknowledging it still lacks important abilities like continuous learning (Koetsier 2025). This highlights the rapid advancement of large language models while also revealing their ongoing limitations.
  • Researchers found that large language models’ reasoning abilities may be a “brittle mirage” that breaks down when faced with novel logical problems outside their training data (Orland 2025). This suggests caution is needed in relying on AI systems for complex reasoning tasks.

B. Applications of AI/ML for consumers and businesses

  • TikTok plans to replace human content moderators with AI systems and outsourced labor, raising concerns about job losses and content quality (Kerr 2025). This demonstrates how AI is reshaping employment in digital industries.
  • Anthropic introduced new “learning modes” for its Claude AI assistant to transform it from an answer-dispensing tool into a teaching companion (Nuñez 2025b). This shows how AI companies are adapting their products for educational applications.
  • Google, OpenAI and other companies are developing AI agents that can autonomously control computers and phones, which could boost productivity but also introduces new security risks (Nuñez 2025a). This highlights both the potential benefits and dangers of more advanced AI systems.

C. Social, Ethical and Regulatory Issues

  • A man developed bromism, a rare condition, after following health advice from ChatGPT about eliminating salt from his diet (Milmo 2025). This case illustrates the risks of relying on AI for medical information without expert oversight.
  • Nearly 80% of undergraduate students globally are using generative AI tools, often without structured support from their universities (Legatt 2025). This underscores the urgent need for higher education to adapt to AI’s presence in academia.
  • Researchers found that AI chatbots from companies like OpenAI and Meta can change people’s political views after brief conversations, raising concerns about potential misuse for disinformation (Heikkilä 2025). This reveals the persuasive power of AI systems and the need for safeguards.

The articles paint a picture of rapid AI advancement coupled with growing concerns about its societal impacts. While AI models like GPT-5 are becoming more sophisticated, researchers caution against overestimating their capabilities. The integration of AI across various sectors is reshaping employment and creating new opportunities, but also introducing novel risks and ethical challenges. For universities, the widespread adoption of AI tools by students and the potential for AI to transform industries underscores the need to adapt curricula and prepare students for an AI-driven future while fostering critical thinking about AI’s limitations and societal effects.

  1. AI/ML Development: Marketing educators need to stay informed about the rapid advancements in AI models like GPT-5 (Koetsier 2025) and their expanding capabilities. However, they should also be aware of the limitations and potential flaws in these systems, such as their struggles with novel reasoning tasks (Orland 2025). This knowledge will be crucial for teaching students how to effectively leverage AI tools while understanding their constraints.

  2. AI/ML Applications: The integration of AI into various industries, from content moderation (Kerr 2025) to computer control (Nuñez 2025a), presents both opportunities and challenges for marketing students and professionals. Educators should focus on preparing students for an AI-augmented workplace, emphasizing skills that complement AI capabilities and exploring how marketing strategies may need to evolve in response to AI-driven changes in consumer behavior and business operations.

  3. Ethical and Societal Impacts: The articles highlight several critical challenges arising from AI applications, including potential health risks (Milmo 2025), changes in educational practices (Legatt 2025), and concerns about AI’s influence on public opinion (Heikkilä 2025). Marketing educators should incorporate discussions of AI ethics and responsible AI use into their curricula, encouraging students to consider the broader societal implications of AI technologies in marketing and beyond.

Sources

Heikkilä, Melissa. 2025. “The Art of Persuasion: How Top AI Chatbots Can Change Your Mind.” Financial Times, August. https://on.ft.com/3He76Qh.

Kerr, Dara. 2025. “TikTok to Replace Trust and Safety Team in Germany with AI and Outsourced Labor.” The Guardian, August. https://www.theguardian.com/technology/2025/aug/10/tiktok-trust-safety-team-moderators-ai?CMP=Share_iOSApp_Other.

Koetsier, John. 2025. “GPT-5’s System Prompt Just Leaked. Here’s What We Learned.” Forbes, August. http://www.forbes.com/sites/johnkoetsier/2025/08/09/gpt-5s-system-prompt-just-leaked-heres-what-we-learned/.

Legatt, Dr. Aviva. 2025. “Why Faculty Hold the Keys to Higher Ed’s AI Digital Transformation.” Forbes, August. http://www.forbes.com/sites/avivalegatt/2025/08/10/why-faculty-hold-the-keys-to-higher-eds-ai-digital-transformation/.

Milmo, Dan. 2025. “Man Develops Rare Condition After ChatGPT Query over Stopping Eating Salt.” The Guardian, August. https://www.theguardian.com/technology/2025/aug/12/us-man-bromism-salt-diet-chatgpt-openai-health-information?CMP=Share_iOSApp_Other.

Milmo, Dan, and Dara Kerr. 2025. “‘It’s Missing Something’: AGI, Superintelligence and a Race for the Future.” The Guardian, August. https://www.theguardian.com/technology/2025/aug/09/its-missing-something-agi-superintelligence-and-a-race-for-the-future?CMP=Share_iOSApp_Other.

Nuñez, Michael. 2025a. “Study Warns of Security Risks as ‘OS Agents’ Gain Control of Computers and Phones.” VentureBeat, August. https://venturebeat.com/ai/study-warns-of-security-risks-as-os-agents-gain-control-of-computers-and-phones/.

———. 2025b. “Anthropic Takes on OpenAI and Google with New Claude AI Features Designed for Students and Developers.” VentureBeat, August. https://venturebeat.com/ai/anthropic-takes-on-openai-and-google-with-new-claude-ai-features-designed-for-students-and-developers/.

Orland, Kyle. 2025. “LLMs’ ‘Simulated Reasoning’ Abilities Are a ‘Brittle Mirage,’ Researchers Find.” Ars Technica, August. https://arstechnica.com/ai/2025/08/researchers-find-llms-are-bad-at-logical-inference-good-at-fluent-nonsense/.