How comfortable are you with sharing your medical history with an AI? I’m certainly not.

OpenAI’s announcement about its new ChatGPT Health program prompted discussions about data privacy and how the company plans to keep the information users submit safe.

ChatGPT Health is a dedicated “health space” inside ChatGPT that lets users connect their medical records and wellness apps so the model can answer health and wellness questions in a more personalized way.

OpenAI promises additional, layered protections designed specifically for health, “to keep health conversations protected and compartmentalized.”

First off, it’s important to understand that this is not a diagnostic or treatment system. It’s framed as a support tool to help understand health information and prepare for care.

But this is the part that raised questions and concerns:

“You can securely connect medical records and wellness apps to ground conversations in your own health information, so responses are more relevant and useful to you.”

In other words, ChatGPT Health lets you link medical records and apps such as Apple Health, MyFitnessPal, and others so the system can explain lab results, track trends (e.g., cholesterol), and help you prepare questions for clinicians or compare insurance options based on your health data.

Given our reservations about the state of AI security in general and chatbots in particular, this is a line that I don’t dare cross. For now, however, I don’t even have the option, since only users with ChatGPT Free, Go, Plus, and Pro plans outside of the European Economic Area, Switzerland, and the United Kingdom can sign up for the waitlist.

Users should realize that health information is very sensitive and as Sara Geoghegan, senior counsel at the Electronic Privacy Information Center told The Record: by sharing their electronic medical records with ChatGPT Health, users in the US could effectively remove the HIPAA protection from those records, which is a serious consideration for anyone sharing medical data.

…

Categories:

Clever hackers and digital privacy advocates are fighting back against the snooping activities of Kristi Noem’s masked agents. The Electronic Frontier Foundation (EFF) has rounded up several of these counter surveillance projects, and perhaps unsurprisingly many of these have to do with Flock, best known for its automated license plate reader (ALPR).

Flock operates the largest network of surveillance cameras in America, and, while it has contracts with thousands of police departments and municipalities across the US, sometimes ICE gains access to this footage, according to US Senator Ron Wyden (D-OR) and those who have looked into Flock’s misuse.

We should also note that EFF and the American Civil Liberties Union (ACLU) have sued the city of San Jose, California, over its alleged abuses of this technology.

One YouTuber discovered a way to prevent your license plate being recorded and logged by Flock’s AI readers by screen printing “tiny bits” of adversarial noise and putting the sticker on your license plate. These “abstract invisible license plate overlay patterns … cannot be detected by humans but make license plate recognition systems utterly shit the bed,” Benn Jordan said on his video.

Jordan also uncovered a massive Flock security snafu involving hundreds of misconfigured Flock cameras that exposed non-password protected admin interfaces to the public internet, allowing anyone to view live surveillance feeds, download videos, and view logs. “Like a Netflix for stalkers,” is how Jordan described it.

Plus, apps including Stop ICE Alerts, ICEOUT.org, and ICEBlock allow users to report local ICE sightings.

…

The Verge – no paywall: “Earlier this month, The Guardian published an investigation that showed Google was serving up misleading and outright false information via its AI overviews in response to certain medical inquiries. Now those results appear to have been removed. According to the original report:

Apple’s decision to add Google Gemini to Siri will also keep users’ personal information safe by making sure Google can’t see it. Apple keeps its security high while greatly improving Siri’s performance by using Private Cloud Compute and stopping Google from using queries to train AI.Apple and Google just announced a landmark deal to bring Gemini AI to Siri. Following the announcement, a question came to mind for many Apple users: Will my data be safe? Well, a recent joint statement seems to answer that. It suggests that Apple will use Gemini‘s AI “brain power” without giving Google control over your data.

Google won’t run Gemini-powered Siri request: Apple’s Private Cloud Compute. The backbone of this agreement is Apple’s Private Cloud Compute (PCC). Even though Google’s Gemini models are providing the underlying intelligence, those models won’t be running on Google’s servers. Instead, they will operate within Apple’s own secure infrastructure.

…

Category: https://www.androidheadlines.com/category/apple

Subject: Nothing Is Secure
Source: Columbia Journalism Review
https://www.cjr.org/news/hannah-natanson-fbi-washington-post-raid-devices-seized-runa-sandvik-security-computer-phone-laptop-sources.php[

Via beSpacific: “Early this morning, federal agents executed a court-approved search warrant at the private residence of Hannah Natanson, a reporter for the Washington Post. The agents searched her home and seized multiple personal electronic devices, including her phone and other digital equipment used for reporting…”

See also via CJR  Nothing Is Secure – The home of Hannah Natanson, a Washington Post reporter, was searched by the FBI. Her devices were seized. Runa Sandvik, whose life’s work is protecting journalists’ digital security, assesses the damage—and what news organizations need to know.