In a recent The Conversation article, a duo of Harvard University professors, Bruce Schneier and Nathan Sanders, highlight how many large companies and businesses are incorporating AI into their operations—including advertising—raising concerns about potential manipulation.
They argue that as the AI industry is shifting towards monetizing consumer attention, similar to social media. OpenAI’s launch of ChatGPT Search and Atlas has sparked a race to capture online data for advertising, despite concerns about trust and user skepticism.
Schneier and Sanders point out that AI-driven search companies, including Perplexity, Microsoft Copilot, Google (AI Mode for search), and Amazon (Rufus), have began adding ads ito their platforms. Recently, OpenAI announced it will start test ads in the unpaid version of ChatGPT. Security experts are raising concerns that AI companies might manipulate user behavior for profit, which could have serious implications for the future of AI development. The authors are particularly worried about OpenAI’s recent move with its Search and Atlas, as it signals a shift toward an ad-based model, similar to what Google uses for its search engine.
Here are the key points from Schneier and Sanders’ article:
- Advertising model: The authors explain that Google, which was declared a monopolist by a U.S. federal court, generates most of its revenue from advertising—about 80-90% of its total income. Despite its market dominance, Google Search has become less user-friendly, often prioritizing low-quality content like spam and paid ads over relevant, AI-powered results.
- Highly persuasive: The authors highlight that paid advertising in AI search can subtly influence users’ thoughts, spending habits, and beliefs through direct interaction. This raises concerns about bias and persuasion. Research shows that AI models can be just as effective as humans in shaping perceptions and behaviors, which could affect online communication and expression.
- A different way forward: The authors argue that today’s AI is fundamentally untrustworthy, not because of the technology itself, but because of corporate interests. They suggest that governments could protect consumers by regulating data control, similar to the EU’s approach, creating data protection agencies, and investing in Public AI model. Consumers, they say, are demanding transparency, privacy, reliability, and security from AI, and rejecting its exploitation.
In conclusion, Schneier and Sanders emphasize the need for stronger regulation and transparency in AI to protect consumers from manipulation and ensure technology serves the public good.
Acknowledgement: This summary was generated using AI writing tools as part of my experimental learning process. For more detail, check out the original The Conversation article below.