In the news...
- A.I. Chatbots Want Your Health Records. Tread Carefully. [nytimes.com]
Big Tech wants access to your most sensitive data yet: your health records. Microsoft just launched a Copilot Health tool that pulls together medical records and fitness tracker data to give users AI-generated health summaries, and Amazon, OpenAI, and Anthropic are all testing similar products.
The privacy risks are real: unlike traditional healthcare providers, tech companies aren't bound by HIPAA, meaning they could theoretically use your health data to train AI or serve targeted ads. And the reliability concerns are just as worrying; researchers found that AI chatbots are no better than a Google search at guiding users to correct diagnoses, and one man ended up in a psychiatric unit after following ChatGPT's advice.
- As Meta Removes Privacy Controls, TikTok Explains Why It Never Had Any [fortune.com]
Starting May 8, Instagram will have access to your DMs again. "Meta is ending support for end-to-end encrypted direct messages — reversing a feature it introduced just two years ago — and reopening the door to automated content scanning, AI-powered moderation, and easier compliance with law enforcement requests. TikTok, meanwhile, confirmed it never offered the protection at all. Together, the moves signal that the era of unconditional privacy promises on social media is over.” If privacy is your priority, there are designated end-to-end encrypted messaging apps you can try instead!
- I'm a Cybersecurity Professional, Here's Why I'm Preparing for an AI Data Breach [techradar.com]
Your AI tools might know more about you than you'd like...and that's exactly the problem. Mike Kosak, Director of Threat Intelligence at LastPass, argues that a major breach of an AI company is basically inevitable, given how much sensitive data people and companies are feeding into these tools. One study found that 99% of organizations already have sensitive data exposed to AI tools, and that's before you factor in people using chatbots for deeply personal conversations like mental health support.
Kosak's advice: think carefully about what you share with AI platforms now, before a breach makes that decision for you. (Our advice: You can also use Duck.ai to chat more privately with Claude, ChatGPT, and other AI chatbots. Duck.ai chats are anonymized by DuckDuckGo, with no account necessary, and your data is never used to train the models.)
Proudly Private,

Dax the Duck
Mascot - DuckDuckGo
P.S. Our newsletter doesn't track you, but about 85% of other emails do!
Get a @ duck dot com forwarding address to remove trackers and protect your email address.
Learn more.
Follow us on X
Learn about what’s new at DuckDuckGo on our blog
Go behind the scenes with the DuckDuckGo team on Substack
Join our remote team! See open positions
|