Revisiting “front page of the newspaper” wisdom

I’ve been preparing for my presentation at the Advertising and Marketing Law Conference on 15 October and reading through some materials I’ll probably reference in my slides. One paragraph just stood out for me in Anil Dash’s article “What is Public?“:

The conventional wisdom is “Don’t publish anything on social media that you wouldn’t want to see on the front page of the newspaper.” But this is an absurd and impossible standard. The same tools are being used for person-to-person conversations and for making grand pronouncements to the world, often by the same person at different times. Would we say “Don’t write anything in a sealed letter that you don’t want to see on the front page of the newspaper” simply because the technology exists to read that letter without opening it?

I think the reason this stood out for me is because conventional wisdom is that you shouldn’t publish anything online that you wouldn’t want published on the front page of a newspaper or on a billboard at a busy intersection. It makes sense until you consider that we are using the same platforms to share things privately and publicly.

How many people use Twitter for personal sharing as if they and their Twitter friends are the only people who can see otherwise public updates? They certainly don’t intend for their tweets to be shared with everyone who uses Twitter (until they do) and although Twitter is very public (unless you lock down your profile) many of its users still have this illogical expectation that their tweets are not for public consumption.

If anything, this sort of issue highlights how complex privacy is in this digital age. We face a number of tough questions about how we use social media and what seemingly obvious notions like privacy really mean to us.

Apple tells developers not to share health data with advertisers

According to The Guardian, Apple has imposed contractual restrictions on developers that prohibit them from sharing health data they may receive through an anticipated range of health-related apps which iOS 8 will usher in through a platform called HealthKit:

Its new rules clarify that developers who build apps that tap into HealthKit, of which Nike is rumoured to be one, can collect the data it holds.

But, they stated, the developers “must not sell an end-user’s health information collected through the HealthKit APIs to advertising platforms, data brokers or information resellers”. Although, the rules add that they could share their data with “third parties for medical research purposes” as long as they get users’ consent.

These sorts of apps have enormous potential to benefit consumers and, at the same time, they represent a profound risk to consumers because our most intimate personal information is being accessed. How developers and device manufacturers handle this data is bound to inform a new generation of privacy complaints and reputational harm case studies in the years to come.

Your future digital government

I had to apply for unabridged birth certificates for our children the other day so I sat down in front of my laptop, browsed to the Department of Home Affairs’ website and logged into the secure Civic Services portal to start the process. I used my new ID card with its embedded personal digital certificate and a one-time code from my smartphone to authenticate myself.

As you can imagine, Home Affairs has all my details and who our kids are so all I really had to do was select the option for the unabridged birth certificates and place the order. The system informed me that because this was the first time I had requested these particular birth certificates there wouldn’t be a charge. I received a confirmation of my request along with digitally signed and locked digital versions of our kids unabridged birth certificates about five minutes after I concluded my request.

The birth certificates were in PDF and I quickly verified that they were signed by Home Affairs using the Department’s current public key (they were) and then forwarded them on to the service provider that requested them from us.

At this point you are probably wondering how I managed to do all of this? You probably had to drive out to your local Home Affairs office, fill out the forms on paper and wait in line to hand the forms over to the person behind the counter and be told you’d have to wait six to eight weeks for the birth certificates to be printed out in Pretoria and delivered to that office. You would then have to return to the office with your receipt so you could collect the pages.

My story is completely hypothetical. That process is not currently possible at the moment. This isn’t because the technology doesn’t exist, it does, or because the law doesn’t currently cater for it, it does. Implementing processes like this requires a different approach to digital government services. In this particular case, the starting point is likely a combination of a number of factors:

  • A secure, complete and accurate citizens’ and residents’ database;
  • A secure portal through which citizens and residents can access government services using a unique digital identity which is linked to the data the government has about them;
  • Digital certificates issued to each citizen and resident along with each person’s national identity;
  • A convenient means of both securing and using a digital identity to authenticate each citizen and resident that has a cross-platform mobile as well as conventional desktop interface.

The Electronic Communications and Transactions Act provides a broad framework for much of what would be required, including digital signatures, digital documents and data retention and evidence. The benefits could be to radically streamline government services and empower citizens to transact more securely and effectively with each other. These benefits are not reserved for government services, they extend to private services too. In fact, a single secure and digital identity for South Africa’s inhabitants could serve as a platform for a variety of providers to develop engagement models that could transform how the country functions.

So why isn’t such a system being developed (or in place already – much of the technology required has probably existed for some time now). The Verge has an interesting post on this topic titled “Our future government will work more like Amazon” which has a few relevant observations, including this one:

The problem is logistics. Sure, the Postal Service would probably love to have some fresh resources to boost up these facilities. But consolidating many offices into one is never easy. And reappropriating human resources would definitely be controversial. But with good digital systems to reduce paperwork, remember previous encounters with citizens, and greatly reduce the need for people to visit brick and mortar offices in the first place, it’s certainly feasible.

From a legal perspective there are very few barriers to this sort of future. Aside from logistics, the challenge is that our culture is still heavily invested in paper and paper paradigms and the change to digital workflows seems to be prohibitively complicated. That said, there are many benefits to going digital including cost savings, better security and improved redundancy (if you work with paper files, how much redundancy is built into your filing system?).

Simply adopting the necessary technologies isn’t going to solve the problem either. Effective implementation is essential and failing to do this has led to controversies such as the SANRAL consumer data exploits we read about recently. I came across another example of poor implementation when I began writing this post this morning, somewhat ironically from the South African Post Office’s Trust Centre which is charged with delivering trusted digital identity solutions:

SAPO_Trust_Centre_screenshot_-_authentication_problem

Leaving aside what must be an oversight, the Trust Centre delivers a key component in this future digital economy. An advanced electronic signature, for example, opens the door to a range of digital transactions otherwise reserved for paper-based transactions. One of the things I would like to do, as an attorney, is commission affidavits digitally. That is only legally possible if both I, as the attorney, and the person who wants to have an affidavit commissioned have advanced electronic signatures. At the moment this has to be done in person but when both parties have advanced electronic signatures (and have been authenticated by the Trust Centre), this could probably take place remotely. That, alone, represents a cost and time saving. Other transactions which become possible include digital contracts to sell land and even truly digital wills.

Going digital can transform how we function and how businesses and government operate. It just takes vision, an understanding of the legalities and risks and sensible technology implementations.

How you could sell a car in 2034 without a hefty contract

This article was originally published on LinkedIn as “How I sold my car in 2034“.

En Vauxhall bliver demonstreret af sælgeren

Prologue

Contracts are increasingly complex and difficult to navigate, even with recent efforts to simplify the language we use. Much of this is the result of efforts to express complex and interrelated legal and compliance concepts in words and since every legal writer has his or her individual style, the variations in contracts are staggering.

Recently one of my clients asked me for a single paragraph that somehow encapsulated a vendor contract. My response was that such a thing is extremely risky and not pragmatic. There is simply too much in a contract of that nature to adequately express in a single paragraph. Instead, I suggested a couple of options that streamlined the interface for the complete contract.

Later that night I thought about the request further and what it would take to create a “1 paragraph contract” for my client. I realised that such a thing would look very different to the contracts we have now. In fact, the path to a an effective contract that could be expressed in such a short form could lead to a radical overhaul of the broader legal and compliance environment that underpins almost everything we do.

Imagine that instead of expressing those complex and interrelated legal and compliance concepts in words, we reverse the process and establish a syntax to express those concepts more abstractly and yet in a way that still includes all that stuff the “fine print” is designed to cater for in our every day dealings? We could develop a new way of going about our business that doesn’t require lawyers writing pages of contracts that may still be susceptible to interpretational differences.

Going further, what if the way we contract ties directly into a broader contracts profile we all have from our first contract and which gives assurances as to what we can legitimately contract for? This is just the beginning of what could be possible. Legal frameworks could be developed, implemented and enforced programmatically. It would mean a radical transformation of the legal profession, possibly the end of much of the profession as we know it today. On the other hand, it would mean that people could go about their lives, dealing with each other with more confidence, far less uncertainty and without needing to spend so much on unintelligible legal fees.

The story below is a hypothetical scenario which should give you an idea of how this could work. Whether this scenario becomes a reality one day is another question altogether. I suspect that two developments will be key drivers: the so-called Internet of Things and cognitive systems like IBM’s Watson.

How I sold my car in 2034

I arranged to meet Andre on a sunny Sunday morning, 28 May 2034, to sell him my vintage car. I hadn’t met him in person before but I knew it was him because I received verification of his identity when we shook hands and sat down through an interaction between our CitIdents, the SmartNet and some or other authentication process my contract technician told me happens in the background. We met in a local teahouse and chatted while the waitron delivered our orders. Andre asked me about my run earlier that morning (my best time yet) and I congratulated him on his daughter’s latest masterpiece which he shared the night before. We then turned our attention to the deal we were about to make.

Andre kicked off the discussion with a quick data request to access the car’s entry in my Registry. He reviewed the car’s purchase and service history along with its logged mileage and general condition. It pretty much matched the data representation I posted with my sale ad the week before and he was also able to confirm that I was the car’s owner and entitled to sell it to him in the first place. He didn’t say anything but I suspect he also ran a quick valuation check through SmartNet to confirm my asking price was reasonable. This sounds like a lot but he finished his initial review in the time it took me to empty a sachet of sweetener into my tea and stir it.

He smiled and said he was comfortable with the car’s history and condition as well as my price. We exchanged data requests for access to the relevant portions of our contract profiles in our respective CitIdents (this has become standard practice when contracting these days). We both received confirmation that we had the necessary legal capacity to sell and buy the car (Andre’s verification included confirming with my bank that I have paid my vehicle finance and the bank had transferred ownership to me). Andre’s bank confirmed with me that he had sufficient funds to pay for the car on our agreed terms and established a payment link to my bank account for a one way funds transfer.

We decided, for the sake of tradition, to conclude our contract with a handshake. Our wrist tokens registered each other’s proximity as I said “I, Paul Jacobson, agree to sell you my car for our agreed price today.”. Andre smiled again and, in return, said “I accept your offer to buy your car today.”. With that our respective CitIdent’s registered the details of our agreement: the car being sold, our agreed purchase price, the current date and time as well as our verified identities. The SmartNet quickly polled our CitIdents for the further information it required to complete the legal and logistical aspects of our deal, advised the relevant local authorities so they could update their records and I received a data notification that the car had been removed from my Registry and transferred into Andre’s along with confirmation of the first of Andre’s payments.

We chatted a little more, finished our tea. Andre took a call from his partner and while he was chatting, I took a moment to review the transaction records newly associated with my CitIdent’s contract profile. Sure enough the sale was symbolically represented using the usual cheerful info-icons with the broad parameters of our transaction supplemented with the usual conditions, restrictions and permissions provided by the SmartNet’s latest contracts AI. The latest models finally introduced cross-jurisdictional compatibility between different regions’ contract models.

Andre finished his call, I sent payment to the waitron with a tip and thanked Andre. For a moment I couldn’t understand why the car didn’t respond to my proximity and unlock and then I realised it wasn’t mine any more. Senior moment. Since it was a lovely Autumn morning I decided to take a pod home and spend the rest of the day with my wife and children.

The legalities of going paper-less

Many people think that going digital is problematic from a legal perspective and, to a large extent, it isn’t. It does require that you understand the implications and adjust your workflows to suit the digital paradigm.

Processed with VSCOcam with h1 preset

First Coast recently asked me to speak to clients and media about the legalities of shifting to a paper-less (or even paper-free) workflow. This is something I am passionate about and I have developed a paper-less over the years so I was excited about this talk.

Many people think that going digital is problematic from a legal perspective and, to a large extent, it isn’t. It does require that you understand the implications and adjust your workflows to suit the digital paradigm.

It turned out that First Coast distributes the Avision scanner range which includes models which are rebranded as Fujitsu ScanSnap scanners. I talked about my ScanSnap scanner in my talk and it is easily one of the most valuable devices in my office to me. Discovering that First Coast distributes Avision scanners was a highlight for me because it means that if my beloved ScanSnap moved on to the Big Office in the Sky one day, local help is at hand in the form of an updated version of this critical device.

I published the slides for my talk to SlideShare just before I spoke and you can find them here:

I received an email from one of First Coast’s media partners, Craig Neill from CraCin Productions, letting me know that the video from my talk is up on YouTube too:

This video looks great if you watch the video at 720p.

How to deal with stalkers taking photos of you

I spoke to Kieno Kammies on 567 CapeTalk radio this morning about a troubling trend. As you can hear from the segment, below, the concern is partly about people being photographed in suspicious ways in public. One example is a person following women around shooting video of them or taking photos without their knowledge. This isn’t so much about a person taking a photograph of a scene that happens to include women walking past but actually targeting those women.

Whether this is a privacy issue depends very much on the subject matter and the context. In this respect it comes down to legitimate expectations of privacy in the case of adults and appropriate consent when it comes to children (at least in terms of the Protection of Personal Information Act). The law that is likely to be more appropriate here is the Protection from Harassment Act which targets forms of harassment which the Act defines as follows:

harassment” means directly or indirectly engaging in conduct that the 5 respondent knows or ought to know-

(a) causes harm or inspires the reasonable belief that harm may be caused to the complainant or a related person by unreasonably-

(i) following, watching, pursuing or accosting of the complainant or a related person, or loitering outside of or near the building or place where the complainant or a related person resides, works, carries on business, studies or happens to be;

(ii) engaging in verbal, electronic or any other communication aimed at the complainant or a re.lated person, by any means, whether or not conversation ensues; or

(iii) sending, delivering or causing the delivery of letters, telegrams, packages, facsimiles, electronic mail or other objects to the complainant or a related person or leaving them where they will be found by, given to. or brought to the attention of, the complainant or a related person; or

(b) amounts to sexual harassment of the complainant or a related person;

<

p>The harm the Act protects against may be “any mental, psychological, physical or economic harm”.

This Act is designed to be user friendly and the Regulations describe which forms to use for which steps and who to approach at each step. The Department of Justice and Constitutional Development has a comprehensive page with links to the Act, the Regulations and the various forms. The process was designed in such a way that you don’t need an attorney to assist you (although you can have one helping you) and you need not know the harasser’s identity either. The Act creates a mechanism whereby the police may be instructed to investigate and identify the suspected harasser.

This legislation can be used for a various activities which fall into the “harassment” definition including stalkers like the ones described in the segment as well as cyber-bullying and more.

Are banks assessing your creditworthiness based on your social media profiles?

GigaOm has an interesting article titled “New breed of lenders use Facebook and Twitter data to judge borrowers” which looks at a growing trend in financial services industries. Banks and other lenders are starting to look at customers’ social media profiles when assessing their needs and the risks they may pose as debtors. An emerging South African consumer protection framework could support extension of this behaviour to South Africa, if it hasn’t already been adopted.

GigaOm has an interesting article titled “New breed of lenders use Facebook and Twitter data to judge borrowers” which looks at a growing trend in financial services industries. Banks and other lenders are starting to look at customers’ social media profiles when assessing their needs and the risks they may pose as debtors. An emerging South African consumer protection framework could support extension of this behaviour to South Africa, if it hasn’t already been adopted.

At first glance this may seem somewhat outrageous (and it reminds me of the prospect of insurers using similar data to assess and modify premiums for insurance cover) but this sort of behaviour could well be supported by new consumer protection frameworks being adopted in South Africa.

The Treating Customers Fairly framework is in the process of being adopted by South African financial institutions:

is an outcomes based regulatory and supervisory approach designed to ensure that specific, clearly articulated fairness outcomes for financial services consumers are delivered by regulated financial firms.

Participating firms are required to show they deliver six outcomes which include these two:

  • Products & services marketed and sold in the retail market are designed to meet the needs of identified customer groups and are targeted accordingly; and
  • Where advice is given, it is suitable and takes account of customer circumstances

These two criteria could well lend support for a similar practice of analysing customers’ social media profiles and streams when assessing customers’ needs and risks under the guise of improved compliance with Treating Customers Fairly.

NSA spying negates attorney-client privilege

Revelations about US and UK intelligence agencies’ global Internet surveillance campaigns have highlighted the need for lawyers to take further steps to secure client data using more robust encryption technologies.

To say recent revelations about the US National Security Agency’s global surveillance campaign are unsettling is an understatement. While we are still getting to terms with the extent of the surveillance we have all been subject to, a few things are pretty clear:

  • The NSA has been collecting both the metadata relating to our digital communications as well as their content (the US government has maintained that it does not collect the content of US citizens’ communications but that has been challenged).
  • Internet services we rely on are not immune to the NSA’s campaign. On the contrary, their data is being accessed, either directly or indirectly, and stored for future reference in case they do something illegal or problematic in the future.

You may think that if you don’t rely on the larger Web service providers (apparently with the exception of Twitter which, as I understand it, refuses to co-operate with the NSA) you are immune from scrutiny but that is a fallacy. The NSA has basically tied itself into the Internet’s backbone and monitors all traffic passing through the routers and infrastucture it can see. What this means is that if the data you send or receive passes through any of the Web services that have been co-opted by the NSA under the USA’s secrecy laws or the very Internet infrastructure the NSA is monitoring, your data is exposed to the NSA’s scrutiny.

What we don’t know is just how the NSA analyses the data and to what extent. Certainly as non-US citizens, we are being targeted and all our communications are potentially being captured and stored for analysis. The Guardian has also reported that this isn’t limited to the NSA. The British GCHQ (the equivalent of the NSA) is also monitoring data passing through the UK. We already know that the South African government has the mechanisms in place to monitor our communications through legislation like the Regulation of Interception of Communications and Provision of Communication-related Information Act which essentially mandates storage of users’ identities and their mobile device data. It is conceivable that our government’s programs are more extensive. The NSA’s and GCHQ’s programs certainly seem to grant other nations implicit permission to run similar programs.

Aside from the dramatic erosion of privacy, one significant concern is how these programs impact legal professional privilege (which incorporates attorney-client privilege). The Electronic Freedom Foundation recently published a detailed review of what has come to light about the NSA’s surveillance programs and dealt specifically with attorney-client privilege:

Attorney-Client Privilege Means Nothing

The attorney client privilege is a long-standing feature of American law, one of the oldest and most cherished privileges through out the ages. As one court explained, it is the cornerstone of the privilege is “that one who seeks advice or aid from a lawyer should be completely free of any fear that his secrets will be uncovered.”

The NSA document shows they cut through this privilege like a hot knife through butter. The NSA only has to stop looking at the communication if the person is known to be under criminal indictment in the United States and communicating with her attorney for that particular matter.

This remarkably myopic view of the privilege means communications between attorneys and clients in many cases will be unduly spied on. This is exactly what the ACLU was worried about when they challenged the constitutionality of the FISA Amendments Act. They alleged that attorneys working with clients overseas had an ethical obligation not to electronically communicate with them because the NSA was likely able to read their emails. While the Supreme Court dismissed their suit for lack of standing, these documents at least in part, confirm their fears.

This could also mean any attorney-client communications with someone like Julian Assange of WikiLeaks, who has never been publicly acknowledged as indicted in the U.S., would be fair game.

Even where the privilege applies, the NSA does not destroy the information. The privileged nature is noted in the log, to “protect it” from use in criminal prosecutions, but the NSA is free to retain and use the information for other purposes. No limits on other uses, so long as the NSA General Counsel approves. This is a complete perversion of the attorney-client privilege. The privilege is designed to allow free communication of attorneys and those who they represent, so the client can get good counsel without hiding the truth from his attorney. It is not simply about preventing that communication from being used as evidence in a criminal case.

Legal professional privilege, at least in our law, is actually the client’s privilege that binds the client’s lawyers and protects communications between the client and his or her lawyers from disclosure. The Constitutional Court dealt with the legal professional privilege at common law (the Court wasn’t asked to consider privilege as a Constitutional right) in the 2008 Thint (Pty) Ltd v National Director of Public Prosecutions and Others, Zuma and Another v National Director of Public Prosecutions and Others case:

The right to legal professional privilege is a general rule of our common law which states that communications between a legal advisor and his or her client are protected from disclosure, provided that certain requirements are met. The rationale of this right has changed over time. It is now generally accepted that these communications should be protected in order to facilitate the proper functioning of an adversarial system of justice, because it encourages full and frank disclosure between advisors and clients. This, in turn, promotes fairness in litigation. In the context of criminal proceedings, moreover, the right to have privileged communications with a lawyer protected is necessary to uphold the right to a fair trial in terms of section 35 of the Constitution, and for that reason it is to be taken very seriously indeed.

Accordingly, privileged materials may not be admitted as evidence without consent. Nor may they be seized under a search warrant. They need not be disclosed during the discovery process. The person in whom the right vests may not be obliged to testify about the content of the privileged material. It should, however, be emphasised that the common-law right to legal professional privilege must be claimed by the right-holder or by the right-holder’s legal representative. The right is not absolute; it may, depending upon the facts of a specific case, be outweighed by countervailing considerations.

The general principle that lawyers not disclose (or permit to be disclosed) confidential information is practically universally accepted as an essential component of legal professional privilege (or its equivalents in various jurisdictions). The Law Society of South Africa’s Information Security Guidelines quotes what appears to be an early version of the International Bar Association’s International Code of Ethics governing lawyers’ conduct as an example of the expectations of lawyers. Rule 4, in particular, states the following:

Rule 4. All communications between attorney and client relating to the subject matter of the lawyer’s representation are privileged and may not be disclosed without the client’s express or implicit permission, except to the extent they relate to future conduct that may be criminal or fraudulent. In-house attorneys are included within the scope of this Rule.

Rule 6 goes on to state the following:

Rule 6. Lawyers should never disclose, unless lawfully ordered to do so by a proper Court with jurisdiction or as required by Statute or in a lawsuit with the client, what has been communicated to them in their capacity as lawyers even after they have ceased to be the client’s counsel. This duty extends to their partners, to junior lawyers assisting them and to their employees.

Complying with information security requirements generally involves exercising due diligence in assessing a service’s security features and taking reasonable steps to secure client data. Storing client data on a centrally located server within a firm’s offices may be relatively secure from the perspective that the server may not be susceptible to scrutiny from external parties but the days of an offline server a an effective resource are likely over as lawyers are increasingly expected to work flexibly and remotely. It also isn’t always feasible to manage servers and their security requirements in-house, particularly for smaller firms so lawyers look to cloud services for flexible, hosted and secure solutions.

In a time where foreign agencies weren’t conducting indiscriminate and pervasive surveillance campaigns (you have to wonder when there was such a time, in retrospect), it was probably sufficient to ensure that the cloud service was adequate encryption (both for data being transmitted as well as being stored) and sufficient physical access control policies to protect the data from foreseeable attacks. That doesn’t seem to be enough anymore and lawyers are going to have to beef up their security protocols to meet their ethical and legal confidentiality requirements.

What this likely means is that lawyers are going to need to seriously consider implementing encryption technologies for email and their cloud services. The NSA whistleblower, Edward Snowden, seems to think that making use of encryption is adequate protection, provided it is properly implemented:

Encryption works. Properly implemented strong crypto systems are one of the few things that you can rely on. Unfortunately, endpoint security is so terrifically weak that NSA can frequently find ways around it.

What this means is that lawyers have to consider their email communications (and data transmission and storage, generally) holistically. Encrypting email requires that recipients also make use of encryption tools to be able to decrypt the email and will also need to make more consistent use of these tools to communicate securely with their lawyers.

Another implication for the NSA/GCHQ revelations is that services which are subject to government surveillance and access are probably not satisfactory options for lawyers anymore, at least not without modification. It seems that lawyers are going to have to integrate data security practices into their day to day workflows and take steps to educate their clients about the need to do likewise. Both parties should begin encrypting (or, at least, digitally signing) their communications to prevent (or, in the case of digitally signed communications, detecting) interceptions and unauthorised data access. Ars Technica has a pretty good guide to popular email encryption options for Mac, Windows and Linux users. This is not a quick fix. It requires a pretty extensive review of how you approach email but we, the legal profession, just don’t seem to have much choice.

The irony is that encrypting your data makes you a bigger target for the NSA. As the EFF points out –

More appallingly, the NSA is allowed to hold onto communications solely because you use encryption. Whether the communication is domestic or foreign, the NSA will hang on to the encrypted message forever, or at least until it is decrypted. And then at least five more years.

The benefit of encrypting your data is that good encryption (presumably) takes considerably more resources to crack and affords clients far better protection of their data. It also means lawyers are more capable of complying with their obligations to protect their clients’ privilege.

Our digital world has been shaken to its core and as legal professionals, we have little choice but to adapt and take data security more seriously. That means rethinking how we communicate with our clients and which services we use to do that. It also means we have to begin sooner rather than later. Each email or file uploaded to a cloud services is more data exposed to scrutiny by agencies that seem to have little regard for legal professional privilege or privacy.

Update (2013-06-24): It’s also worth reading this CNet article titled “How Web mail providers leave door open for NSA surveillance” for perspective on what some of the major Web mail providers are doing (or not) to secure your emails. It turns out that Google is doing more than most.