Re: Apple changed their documentation at my request but it proves they don't care about privacy

On 2026-09-06 17:37, Jon Ribbens wrote:
> On 2026-09-06, Carlos E.R. <[email protected]> wrote:
>> On 2026-09-06 02:06, Jon Ribbens wrote:
>>> Firstly, I do not like crossposting to so many groups. What group
>>> are you actually reading this thread in, so that I can limit the
>>> crossposts please?
>>
>> I understand, but that could be a problem to people that are already
>> reading on a "different" group.
>>
>> Your comments are easier to understand that Arlen (aka Maria) posts. He
>> is not clearly explaining the issues and wants people to read a lot of
>> documentation, instead of just posting an actual summary of the
>> situation with explanations.
>>
>> Ie, what was Apple doing, why is that bad, what have they changed, is
>> that enough and why, what are the actual dangers to people.
>>
>> Complete, and short text.
> 
> Ok, well to summarise what I have gathered then, which may or may not
> be Maria's opinion but reflects my opinion at this point: there are two
> completely separate issues here, which are unrelated except that they
> are both to do with WiFi location databases.

Thanks.

> 
> 
> 1. Apple are storing the location of "hidden" WiFi Access Points.
>     (My opinion: low priority.)

Only Apple?

> 
> Multiple organisations are gathering and storing the physical
> co-ordinates of the MAC addresses ("BSSIDs") of WiFi Access Points
> around the world, whenever they are seen transmitting by, e.g. mobile
> phones. (Any WiFi-enabled device can see this information; it does not
> need to be connected to the network in question, nor does it need to
> know its password.)

Yes.

> The purpose of these databases of BSSID locations is to assist devices
> such as mobile phones in locating themselves. Maybe the device is
> indoors and cannot get a GPS signal, but also my understanding is that
> GPS can fix an accurate location much faster if it starts already
> knowing vaguely where on the planet it is.

Yes. And they can give an approximate location without using the GPS chip.

> 
> There is broad agreement that these databases should not include BSSIDs
> which are broadcasting WiFi network names ("SSIDs") which end in the
> string "_nomap".

Right.

> 
> Allegedly: Apple are adhering to this exclusion, but are not also
> excluding BSSIDs which are not broadcasting any SSID at all (i.e.
> "hidden" networks). Other organisations (e.g. Google) do exclude such
> "hidden" networks.

And if it is hidden they would not see if the SSID ends in _nomap.

But is there a consensus that hidden SSIDs should not be listed? In 
writing? Maybe there is such a consensus now.

> 
> My complete speculation: older Apple software, written before they added
> the "_nomap" exclusion, doesn't care at all about the SSID, so just
> reports from the phone to the central database the BSSID and location.
> The database thus has no way of excluding "hidden" networks, without
> excluding all reports from older devices. The "_nomap" exclusion is
> achieved by later software versions reporting the SSID if it is seen,
> and any BSSIDs associated with "_nomap" SSIDs then being blacklisted,
> maybe for 6 months or something like that. The likelihood of any
> particular BSSID *only* being seen by old Apple devices and *never* new
> ones is very low, and hence the "_nomap" exclusion more-or-less works.
> 
> My opinion: this failure to exclude "hidden" networks is unfortunate and
> should be fixed, and maybe Apple are indeed fixing it, but it's possible
> it may take some years to achieve due to the multitude of devices
> running old software.

Right.

Related: What did Arlen (aka Maria) achieve? That Apple agreed to remove 
all hidden and _nomap entries, or that they removed only his entry?


> 2. Apple's API is trivial to abuse.
>     (My opinion: high priority.)
> 
> The Apple API to query their BSSID location database is remarkably
> unrestricted. It has little or no rate limiting, doesn't ask for an API
> key, and reports not only the location of the BSSID but also potentially
> a great many other BSSIDs in the locality. It is not beyond even an
> individual person's ability to get a list of most of the Access Points
> in the world and all their locations.

Aha.

> 
> This makes it very easy for people to abuse this data in various ways,
> from stalkers tracking victims to state actors tracking military
> targets.

Well, only of the limited subset of people that carry their AP when they 
move.


> Google's API by contrast essentially reverses the process, and instead
> of the phone asking "Where is BSSID <x>?", it says "I can see BSSIDs
> <x>, <y>, and <z>, where am I?".

Ah. Yes, I can see this is better.

> 
> My opinion: the lack of restrictions on the Apple API is unacceptable,
> and if Apple are unwilling to do something about it then governments
> should pass laws (or enforce existing laws) to make them to do so.
> 

Ok.

> My opinion: it would be difficult for Apple to rapidly switch completely
> to using an API more similar to Google's API, since this would remove
> functionality from devices running old software. However they could
> certainly aim to do this eventually, and there are steps they could take
> immediately to improve the situation (e.g. rate limiting, and not
> providing so many additional answers when asked about an individual
> BSSID). I cannot see any obvious reason why they couldn't make
> significant improvements almost immediately.

Ok, yes.

Thank you for the explanation. Easy to understand, and I agree with your 
conclusions.

-- 
Cheers, Carlos.
ES🇪🇸, EU🇪🇺;