This week we take a closer look at the rapid spread of customer‑facing chatbots on websites and explore the growing risks that come with them.
While they often look like a simple website widget, chatbots are increasingly deeply integrated AI agents with access to sensitive systems, personal data and business processes.

We talk about some of the real‑world snafus that have occurred, and then step through the practical risk and governance considerations for privacy, security and AI teams.

Links:

Article about Air Canada chatbot misinformation case (BBC) ⁠https://www.bbc.com/travel/article/20240222-air-canada-chatbot-misinformation-what-travellers-should-know

Article about swearing chatbot (BBC) ⁠https://www.bbc.com/news/technology-68025677⁠

Article about NYC chatbot telling businesses to break the law (The Markup) ⁠https://themarkup.org/artificial-intelligence/2024/03/29/nycs-ai-chatbot-tells-businesses-to-break-the-law

Article about California fire agency chatbot (The Markup) ⁠https://themarkup.org/artificial-intelligence/2025/07/09/californias-fire-protection-agency-made-an-ai-chatbot-dont-ask-it-about-evacuation-orders

Article about McDonald’s chatbot breach (News) ⁠https://www.news.com.au/technology/online/hacking/dystopian-mcdonalds-ai-chat-bot-olivia-hacked-with-simple-password/news-story/2c0f09f9fb0396f6ca17c56419133fee

Article about Salesforce hacks (AFR) ⁠https://www.afr.com/technology/australian-boards-caught-up-in-global-hack-after-portal-breach-20251031-p5n6ui

AI As normal technology (Arvind Narayanan and Sayash Kapoor) https://www.normaltech.ai/p/ai-as-normal-technology

Credits:

Editing and post-production by Martin Franklin (East Coast Studio) www.eastcoaststudio.com.au