This speech was delivered to the Graduate School of Public and International Affairs at the University of Ottawa on October 21, 2014.

At the risk of stating the obvious, and merely to frame the discussion, my starting point today is this: social, technological and economic developments are transforming the risks and modalities of privacy more rapidly than we have time to assess their ethical implications.

Yet, ethical questions abound. Individuals debate whether the hacking of celebrities’ photos is an egregious violation of privacy or sharing them on the Internet was a relinquishment of that right. In the face of unprecedented surveillance capacity and threats, governments and citizens debate the need for increased collection of personal information in the public interest and its privacy implications. Corporations and consumers are wrestling with the ethical issues of a business model that provides free services and applications in return for the use of personal information for advertising.    

Behind these ethical challenges lies the fact that, in essence, the right to privacy is challenged at its core.  It has moved from the physical realm that we easily master - curtains, locks, whispers…-, to the virtual realm, that for most of us is beyond comprehension – secure passwords, encryption, firewalls. Thus, the right to privacy, or the right to control what others know about us, escapes us. Moreover, it is ethically challenged with claims conflicting with the claim to privacy.

Calls for modernized privacy legislation are heard around the world with relative success. By way of example, India is working on a new privacy bill, the European Regulation on Data Protection is making its way through the European Parliament and Canada has introduced Bill S-4, to adopt the Digital Privacy Act. 

But legislative change may be missing a step. The ethical underpinnings of the right to privacy, and moral foundations of privacy law, should be redefined in accordance with an unprecedented context of risks and modalities of exercise before we legislate on it. At the very least, a modernized ethical framework will serve as a guide in interpreting existing law in an unprecedented context of application. 

Today, I will explore eight current claims posited in conflict with the right to privacy, namely:

Claim #1: The democratic right to know supersedes the right to privacy.

Claim #2: The principle of judicial transparency must prevail over privacy even with the differential privacy impact of the Internet.

Claim #3: The right to privacy cannot restrict the right to freedom of expression. 

Claim #4:  Information is no longer private on a public platform.

Claim #5: The right to privacy must give in to the right to be safe. 

Claim #6: If the Internet is free, consumers should expect the use of their personal information to generate advertising revenues. 

Claim #7: If mining accumulated personal information serves public interest, the individual right to privacy should not impede it.

Claim #8: The duty of accountability for our actions supersedes our right to privacy.   

Before we address these evolving questions, it may be useful to ground the discussion in the established, immutable ethical pillars of privacy. 

1. The basic privacy ethical framework – what and why privacy?

In R v Duarte, [1990] 1 S.C.R. 30, Justice Dickson defines privacy as “ the right of the individual to determine when, how and to what extent he or she will release personal information. ” 

Broadly, Canadian law (the Privacy Act and the Protection of Personal Information and Electronic Documents Act) defines personal information as “information about an identifiable individual.”

In addressing the question “Why privacy”, Daniel Solove , Professor of Law at George Washington University Law School, conveniently breaks the answer down into 10 reasons: 

1. Privacy limits the power others have over us by limiting the information they have about us.

2. Privacy allows respect for individuals in what they choose to disclose about themselves.

3. Privacy allows reputation management and therefore protecting our place in society.

4. Privacy maintains appropriate social boundaries by affording a space of solitude and freedom.

5. Privacy allows trust in the disclosure of personal information through confidentiality.

6. Privacy is essential to autonomy and control over one’s life.

7. Privacy allows freedom of thought and speech.

8. Privacy protects freedom of social and political affiliations.

9. Privacy preserves the ability to change and have a second chance.

10. Privacy gives us a space where we do not have to justify or explain ourselves. 

In one sentence, I would say privacy serves personal integrity through the protection of dignity and autonomy. As all fundamental rights, the test for its legitimate restriction is one of proportionality to public interest in a free and democratic society. To ground the application of this definition to the current context of privacy challenges, it may be useful to identify the main pillars of the ethical framework for privacy:

• Privacy is a visceral need and a fundamental right entailing that personal information cannot be collected without consent unless out of necessity and even then in a manner proportionate to that necessity. 
• Even with consent, modalities of collection, use and disclosure must respect user control.
• User control dictates full transparency in collection, use and disclosure.
• Collection, use and disclosure are limited to what is authorized.
• Information must be secure.
• Individuals must be assured accuracy and recourse.   

I would venture that these ethical parameters are well established. They underpin the ethical and legal  frameworks around the right to privacy. My purpose here is to expand upon the social, economic and technological developments that generate conflicting claims to this ethical premise.  

2. Privacy and the right to know – where do you draw the line?

In the summer of 2014, an already embattled, suspended Canadian Senator was confronted with a paternity suit in addition to the pending charges of fraud. As journalists pounded on this “sure sell” piece of news, others asked “When it comes to a politician’s personal life, where do you draw the line between the right to know and the right to privacy”?

The same question arose in France, in early 2014, when the President was caught on a scooter with only one bodyguard, visiting his mistress. He already had a live-in partner at his official residence, with her own role and publicly funded staff. The media had a field day and he claimed breach of privacy. Again, the question was raised as to the relationship between the right to know, in a democracy, and the right to privacy.  

Ethically, the proportionality rule applies and calls for a test of relevance. The right to control one’s personal information is a fundamental right but the right to know in order to hold politicians accountable is central to democracy. The question in each case is whether the intrusion upon privacy is proportionate to the public interest served.  

In the case of the French President, his detractors were successful in arguing that it was relevant for French citizens to know, to hold the President accountable in his duties: he had made himself physically vulnerable by travelling without sufficient protection and his personal safety was a matter of public interest; public expenditures were allowed for his live-in partner on the basis that, while not married, she was his love partner – the existence of another mistress was putting into question the relationship that served to legitimize public expenses; the existence of a secret relationship  was making him vulnerable to blackmail and therefore jeopardized national interests. 

In the case of the Canadian Senator the question of relevance takes me to a different position: the Senator was suspended and therefore not exercising any duty to be held accountable for; the personal information had no bearing on his duties to Canadian citizens. 

In sum, in a democratic context, the privacy of public office holders is limited. It is limited by the right of citizens to know what is relevant to the exercise of public functions, to hold public office holders accountable. 

3. Judicial transparency on the Internet – can it withstand privacy implications?

Judicial transparency is a fundamental tenet of any democracy. It brings the courts under public scrutiny to ensure fairness and integrity of the decision process. 

The Internet, however, has distorted the impact of judicial transparency. With the posting of decisions, scrutiny has moved from the tribunal to the parties. While it may be justified by public interest in certain cases to publish personal information in a judicial decision, in most cases it constitutes a disclosure of personal information beyond what is necessary and proportionate in the public interest. And therein lies the ethical challenge: should the differential impact of the Internet (wide and permanent dissemination) compared to the physical obscurity of the times when the principle of transparency was established (closed court rooms, paper files) have an impact on the implementation of the judicial transparency principle? 

Again, the established ethical parameters of privacy serve as guides: is it necessary and proportionate, in order to hold the court accountable, to disclose the identifying information of the parties? And doing so on a platform that is globally accessible, well beyond the society in which the decision matters, and permanently so, as Internet information is?

In Canada, when I was leading the Office of the Privacy Commissioner (OPC), we issued guidelines for federal administrative tribunals to uphold the right to privacy while adhering to the principle of judicial transparency: in the same fashion as the European Court of Justice in relation to the right to be forgotten, we recommended that identifying information be removed from decisions posted on the Internet, unless it was necessary and proportionate in the public interest to do so. 

Europe is wrestling with that differential impact of the Internet through claims to “the right to be forgotten”. Earlier this year, the Court of Justice of the European Union was seized of this claim in relation to a Spanish individual who required that legal information about his bankruptcy be removed from the Internet. The information was ten years old. The individual claimed it was irrelevant and should be removed in compliance with the right to be forgotten. The European Court agreed under certain conditions: the right to be forgotten applies only in relation to information that is inaccurate, inadequate, irrelevant and excessive. It entails the right to ask for removal of the information. The Court further ventured in the ethical field adding that economic interests of Internet host companies could not justify the violation of data protection and that the right to be forgotten had to be balanced with freedom of expression and the media.  

The decision of the European Court of Justice reflects, in my view, the pressure of a new technological and economic context on established privacy principles. The decision arcs back to the ethical principles of proportionality and necessity as well as to the right to accuracy. In this case, the intrusion upon privacy was – at least in the European Court of Justice’s view – unnecessary and disproportionate in a context of wide and permanent dissemination, no matter the fact that it proceeded from a legal bankruptcy process.  The story is, of course, not over. Where to draw the line is one that we all as global citizens will be involved in setting through social norms, a richer ethical discussion, and ultimately, perhaps, political debate on the proportionality of judicial openness and publication of one’s personal information.

4. Can the right to privacy restrict freedom of expression?

The limits of the right to privacy in relation to freedom of expression were the subject of ethical and legal debate recently in Canada. The Supreme Court of Canada addressed the issue in the case of Alberta Information and Privacy Commissioner (AIPC) vs United Food Commercial Workers (S.C.C. 2013) and in the discussion around Bill C-13, An Act to Amend the Criminal Code, the Canada Evidence Act, the Competition Act and the Mutual Legal Assistance in criminal Matters Act, also known as the Cyberbullying Bill.

In the Supreme Court of Canada decision, the right to freedom of expression was pitted against the right to privacy in relation to the Internet posting, by a union, of photographs of employees crossing the picket line during a strike. These employees filed a complaint with the AIPC for violation of Alberta’s Personal Information Protection Act  (PIPA). The case made its way to the Supreme Court of Canada. The Court found that the union’s ability to communicate is of fundamental importance in labour disputes and one of labour’s most effective tools. Therefore, it is protected as freedom of expression. The Court found the restriction from posting photographs of employees crossing the picket line to be in violation of freedom of expression, disproportionate to the right to control one’s personal information and unjustified.

This application of the proportionality test may be arguable on the basis of relative harm: posting of individuals’ photographs can be seen as causing greater potential harm to individuals than the restriction of that particular form of communication causes to the union. Nevertheless, the test remains unchallenged: curtailing the right to privacy for freedom of expression is a matter of proportionality to public interest in a free and democratic society.

In relation to C-13, the cyber-bullying bill, which prohibits the non-consensual distribution of intimate images, one of the most articulate arguments on the balance between freedom of expression and privacy is found in the 2014 Submission to Parliament of the Canadian Civil Liberties Association (CCLA) . In short, the CCLA argues that the criminalization of non-consensual distribution of intimate images is overly broad and therefore “may unreasonably violate freedom of expression”. 

The actual text of C-13 creates a new offence for “Everyone who knowingly publishes, distributes, transmits, sells, makes available or advertises an intimate image of a person knowing that the person depicted in the image did not give their consent to that conduct, or being reckless as to whether or not that person gave their consent.” The definition of an intimate image includes photographs, film or video, in which a person is nude or engaging in sexual activity and in respect of which there were expectations of privacy. 

The CCLA puts forward that this new provision criminalizes expression, in violation of the Canadian Charter of Rights and Freedoms. Specifically, the CCLA argues that: i) its application is excessive because it does not require malicious intent, ii) the definition of intimate images is overly broad and the standard of reasonable expectation of privacy will be difficult to implement and iii) that the Bill, which includes as penalties for offenders the prohibition from the Internet and other digital networks, further curtails freedom of expression. 

As articulate as the CCLA’s position is, it is also controversial. Again, on the basis of an ethical framework of proportionality and necessity, many come to the conclusion that the public interest justifies the proposed curtailment of freedom of expression. Once again, we must take into account the differential impact of new information technology: the distribution of harmful information can be massive and permanent; the harm can be overwhelming, particularly taking into account the age and vulnerability of the victims – 50% of teenagers say they have been cyber-bullied, 25% say they have been cyber-bullied repeatedly. Taking these facts into account, many believe the right to privacy of intimate images must be enshrined in law and its violation criminalized to reflect a new technological reality of risk.

Again, the ethical framework remains unchanged: no violation of privacy can be justified unless there is evidence of necessity and proportionality. I do not see proportionality in claiming freedom of expression to allow such an egregious violation of privacy as non-consensual distribution of intimate images.  

5. Is personal information still private on a public platform?

This question has been recently debated in relation to unauthorized posting of intimate images of celebrities. Some commentators argued that once the information is shared via the Internet, it is no longer protected by expectations of privacy. From a cursory look at media reports, the ethical debate seems to lean in favour of privacy: the pictures were not meant to be made public, they were sent to a person of trust and they were obtained through a malicious attack. 

The issue of the private nature of personal information on a public platform has been legally addressed in a 2013 Report of findings of an investigation I led at the OPC. A First Nations activist alleged that two federal government departments were surreptitiously collecting her personal information from her social network activities. Neither department denied it. They argued that the information was no longer private since it was posted on a public platform. 

The OPC took a different view: we found that public availability of personal information on the Internet does not render the information non-personal. Therefore, it cannot not be collected unless for legitimate government business.        

But the private sector seems to be divided on this issue: various surveys show that a majority of employers access social networks to assess candidates, without consent, assuming that public availability of the information removes it from the protection of privacy. They agree that it would be unethical to ask a candidate for the social network password, but it appears that any personal information posted on social networks is seen by employers as a valid source of information on candidates’ character and suitability.

What distinguishes this situation from the OPC report of findings, is that the purpose for the search is more palatable: the candidate has, wittingly or not, made personal information publicly available and employers need it to assess the candidates compatibility with business expectations. Still, ethics of privacy call for respect of personal control over personal information. For social network sites, it means they must offer privacy settings that allow clear user choices and private options. For employers, it means transparency, in providing proper notification. Surreptitious collection of personal information, as was the case in relation to the First Nations activist, goes against the ethics of privacy since it denies personal control over one’s information.

6. Can we have safety and privacy?

The integration of privacy in public safety measures has most recently been raised in the matter of C-13, mentioned above. Specifically, the Bill provides for investigative powers that are seen to be so far reaching to compromise the right to privacy. 

But Bill C-13 is only one example of the ethical dilemmas raised by the convergence of new safety threats and unprecedented surveillance capacity. Other dilemmas include concern about radicalization that may lead to greater collection of personal information on travel and political association of all, to catch but a few; the fear of common crime has brought upon such proliferation of closed circuit television cameras that their use has been normalized; crime on cyber space calls for law enforcement on cyber space, but poses the conundrum of violating privacy (to find the perpetrator) to protect privacy (of the victim); crime prevention initiatives that are based on an integrated multi-agency approach (police, social services, health services, schools) are proven to be most effective but they entail a wider sharing of personal information;  the mobility of people and the globalisation of crime and terrorism requires coordinated intervention between States, therefore international sharing of personal information with increased risks to privacy.

To each dilemma, the same ethical principles lead to the same test: to be legitimate, the encroachment upon privacy to ensure public safety must be necessary, proportionate and effective in that regard, without a less intrusive alternative. And it must be transparent to respect user control, as possible in that context. Legitimacy resides in the answers to these questions: 

• Are the envisaged collection, use and disclosure of personal information necessary to ensure public safety?
• Are they proportionate to that goal? 
• Are they likely to be effective towards that goal?
• Is there no less intrusive alternative?
• To what extent can they and do they respect individual control (for example, notification of surveillance)?
• How transparent are authorities about the collection, use and disclosure of personal information (for example, through legislative debate, public communications)?
• How is collection, use or disclosure limited to what is authorized and is overseen?
• What safeguards are in place to keep the information secure?
• What recourse do individuals have to ensure accuracy of information and legitimacy of collection, use and disclosure?

These questions must be asked with respect to any public safety legislation, policy, programme or measure that entails the collection, use or disclosure of personal information to ensure its legitimacy. 

7. Is personal information the price of free Internet?

The September 13th issue of The Economist puts is squarely under the title “Stalkers Inc”: “Surveillance is the advertising industry’s new business model. Privacy needs better protections”.  

Indeed, one of the most common complaints I hear about violation of privacy relates to online advertising. People call it creepy, feeling stalked as their innocent search for a good holiday spot generates ads for cruises on unrelated sites. Suddenly, they feel someone has their number – and they don’t know who. Yet, the same people enjoy free access to the Internet. The ethical question is:  how far does the delivery of free Internet access allow the use of personal information to finance it? 

The Canadian Marketing Association (CMA) has developed its own Code of Ethics and Standards of Practice that address the right to privacy: in addition to stating the fundamental principles of fair information practices, the CMA imposes on its members to respect user control through the mandatory use of the CMA Do Not Contact Service in case of broad mailings, as well as through mandatory provision of easy opt- out functions; security of information is stated in the list of fair information principles and is further expanded upon in relation to transfer of personal information to a third party; in addition, the CMA guides its members in relation to the ethics of advertising to children and teenagers. 

The OPC also tackled the ethical dilemma of on line behavioural advertising (OBA) as an economic reality to contend with in relation to privacy. Its guidelines on OBA assert respect for user control through the requirement of express consent (opt-in) in relation to sensitive information and implied consent (opt-out) in relation to less sensitive information. Opt-out functions must be easy to use, immediately effective and persistent. Advertisers are still held to the limitations of use as authorized and must destroy or de-identify the information as soon as it is no longer relevant. It is recommended that advertisers avoid tracking children.  

As demonstrated in the OPC Report of Findings on Google OBA that I made public in January 2014, the digital economy is so widely based on the collection and use of personal information that the pressure on privacy must be carefully managed. And the built-in protections, as in the case of Google OBA investigation, are put to the test by the sheer number of ads to monitor for compliance with privacy policies. To me, those were the three  great lessons of the Google OBA investigation: i) personal information is THE currency of the digital economy and therefore its use is inescapable; ii) the operational challenges, even for a company as well-resourced and sophisticated as Google, are very high in monitoring privacy compliance of millions of players placing their ads through auctions within seconds; iii) the legitimacy of collection of personal information to provide ads cannot be currently established for the lack of knowledge of its extent. What is the privacy cost of  free Internet?

In relation to  Google, the company  accepted all of the OPC’s recommendations.

Again, an unprecedented technological and economic context force a return to the fundamental principles of privacy to modernize their implementation: even in relation to free Internet, collection, use and disclosure are subject to legitimate business purposes and must respect user control. The OPC OBA Guidelines offer a path to benefit from the digital economy and respect privacy. Companies must step up with an effective compliance framework to address the inherent privacy challenges. 

8. Can the benefits of data mining justify the collection of personal information?

Data has been called the “new oil” and the example of online behavioural advertising illustrates how personal information has become a currency, sustaining a new economy. But the value of data goes beyond commercial interests. It is also a treasure for governments to govern and for scholars to advance knowledge. Yet, this accumulation and use of personal data raises privacy risks that must be managed in a proper integration of public and private interests. The question is this: can the right to privacy accommodate data mining to further the analysis of collective phenomena such as epidemics, health, social or economic patterns or fraudulent activity? 

This area puts the privacy ethical framework to an unprecedented test. On the one hand, there is benefit for us all to mine the treasure of existing personal data banks. On the other hand, the intrusion and, worse, the acceptance of intrusion upon privacy, diminishes us all by reducing respect for personal dignity and autonomy. Hence, our ethical analysis of data mining must be solidly grounded in the fundamental principles of privacy. 

It is established that personal information can only be collected, used or disclosed without consent only if it is demonstrably necessary and proportionate to that necessity. In relation to data mining this would trigger three main constraints, assuming it is done outside the realistic expectation of obtaining individual consent: i) demonstrating prevailing public interest, ii) ensuring proportionality and iii) respecting user control. Demonstrating public interest is somewhat foreign to the application of privacy law in the private sector. However, if business invokes public interest in its data mining activities, it will have to take on the challenge of proving it.

Demonstrating public interest would rest upon evidence of relevance and necessity of collection and use of the data. Proportionality could be achieved by minimization of intrusion with the highest standards of anonymisation or de-identification of the data. Technologists contend, however, that full anonymisation or de-identification is not possible: personal information can always, somehow be re-constructed.  The proportionality test would then bring into balance the level of de-identification (and therefore protection of individuals) with the collective value of the data mining (the public interest). Concretely, where the re-identification process would be so arduous, and the information revealed non-harmful, risk to privacy may be less than the public interest benefits the data mining yields. The data mining, with proper limitations, would then be proportionate and therefore legitimate.

Perhaps, more importantly, we may need to change our approach to the protection of personal information by moving from a static notion of personal information (defining it on the basis of category – name address, marital status etc…) to develop a dynamic notion of impactful information (non-impactful, slightly impactful, very impactful etc…). We already make the distinction between sensitive and less sensitive information but the notion of “impact” would facilitate weighing public interest and individual impact in a practical rather than theoretical manner. The test would be “if this personal information is collected and used what would be the impact on the individual? And does that impact outweigh the collective benefits from the data mining exercise?” 

Moving to such a concrete test would allow us to ponder the actual benefits of protecting privacy versus allowing data mining. Starting with the premise that every effort would be put towards proportionality and therefore minimization and de-identification, the proportionality of remaining privacy risks could be assessed on a reasonable forecast of impact: for example, how much would it impact an individual if she could be re-identified in a data-mining project on flu trends in Quebec? 

Outweighing of impact though, would not absolve of the duties for data protection. Respect for user control could be ensured in various ways, as operationally feasible: clear notification of data-mining activities, easy opt-out, transparency on purposes, collection, use and disclosure. The collection, use and disclosure must be to what is strictly demonstrated as necessary to the stated purposes and as authorised. The information must be held secure through high-level de-identification and safeguards and individuals must have effective ways of recourse. 

The implementation of an ethical framework for data mining is still in progress and must be developed. There is public interest vested in the “new oil” and it must be realized in full compliance with the fundamental right to privacy, sustained through a clear regulatory framework.  

9. Where does personal accountability stop and privacy start?

For all the media reports and Internet chats on the case of Ray Rice, I have not seen one mention of his or his wife’s right to privacy. In relation to him, no claim to privacy would have stood up to the duty of accountability. In relation to her, no duty of accountability could justify the violation of privacy inherent to the amount and breadth of images shown of her. A proper weighing of the two would have led to greater reserve. 

But reserve seems out of fashion in the digital age and accountability is one concept particularly impacted by the new technological platforms: a rude waiter will be denounced on TripAdvisor, a mean employer will be exposed on a Facebook account and the video of a drunken mayor will go viral. We now have unprecedented parameters of accountability. There are times when the right to privacy and the duty to accountability converge and there are times when they conflict. 

In the case of calls for body-worn cameras (BWCs) for police officers, privacy and accountability may be both accommodated. Civil libertarians argue that the BWCs will keep police officers accountable for their actions and that this prevails over claims of privacy. Police officers claim that BWCs will get the facts straight about incidents and inhibit violence among citizens. Privacy would be ensured through proportionate use, broad notification and security of the information, including limited retention times. 

In addressing the issue in his Independent Review for Toronto Police Chief William Blair, in July 2014, the Honourable Frank Iacobucci brought out the benefits of BWCs – evidence of reduction of violence, improvement of officer and subject conduct, deterrence of false allegations of misconduct, increased transparency-   but also raised significant privacy implications: personal information of officers, subjects and by-standers is collected and may be used; in some contexts, as in entering a home, the use of BWCs by police officers may be overly intrusive; the sharing of the information with third parties is highly consequential. 

The Honourable Iacobucci concludes in favour of the use of BWCs but subject to a strict protocol for the protection of privacy:  the use of BWCs must be transparent and evident to all; it must remain proportionate so activation and de-activation must be possible to avoid unnecessary intrusion upon privacy; information must be destroyed as soon as it is no longer relevant; storage must be secure and clear limitations must be set to use and disclosure. 

While the union of privacy and accountability seems happy in relation to BWCs, it faltered at the altar in the case of a young offender in B.C.

During the 2011 Stanley Cup Vancouver riots, he was photographed setting a police car on fire. He was held accountable through social media with wide distribution of his image, caught in the act. While I would never condone his behaviour, the posting of his picture on the Internet violates a long established principle in Canadian law: he was only 17, thus making him a young offender. The Youth Criminal Justice Act specifically protects the privacy of young offenders through prohibiting public disclosure of their identity or information from which identity can be deduced. This stems from consideration of the particular consequences for youth to be labelled or stigmatised for an act committed at an immature age. Social media have now made this principle inoperative.    

This brings into question current responsibility for compliance with the existing privacy regime as well as the need for legislation to fill the void in relation to individual upon individual violation of privacy. Does accountability for our actions allow others to violate our privacy any time they believe we have transgressed? What is the responsibility of Internet hosts in monitoring the respect for privacy between individuals? This specific question was addressed in several, otherwise unrelated cases.

For example, in the 2009 OPC Report of Findings in an investigation about Facebook, the responsibility of Facebook in relation to privacy protection between individuals was left to what was properly within the company’s control: a timely reminder to users to seek consent before posting someone’s personal information. In 2012, in another investigation of Facebook, this time regarding the harm of an imposter account, the OPC recommended and Facebook agreed to establish a process to help non-users notify individuals that had been “friended” by an imposter. 

In Italy, the specific issue of responsibility of Internet companies was at the centre of a case where three Google executives were charged and convicted in 2010 for violating privacy law. The charges were based on the consideration by the court that Google had allowed the upload of a harmful video of an autistic boy being bullied and once alerted of its existence, was too slow in taking it down. The convictions were overturned and the executives acquitted by the Supreme Court of Italy in 2014, with the following reasoning: companies who serve as hosts of online activity cannot be held responsible for content or asked to review content (the right to freedom of expression was raised in this case as well). Moreover, Google has clear terms of service in relation to the appropriateness of content and took down the video within hours of being notified by police, thus reacting diligently. 

The decision of the European Court of Justice on the right to be forgotten, however, points to greater expectations in relation to online hosts. The right to be forgotten implies the obligation for hosts to remove information that violates privacy. Consequently, as expectations of accountability grow, unbridled, we need to establish boundaries in relation to the right to privacy. Internet host companies will have to be vigilant to ensure respect for privacy on their sites among individuals. Governments will need to address this privacy issue through legislation. 

Conclusion: Grounding new protections in ancient rights

While our information platforms and needs have radically changed, our right to privacy, as a matter of freedom and personal integrity, remains the same. The challenge is to apply the immutable principles of privacy to ever-changing privacy risks and modalities. This requires, in my view, three main qualities:

• Innovative minds to interpret privacy law and apply privacy ethics in a manner that fully integrates privacy protection to evolving social trends, business models and technological platforms. 
• Pragmatism to assert privacy protective measures that are operationally relevant and favour rather than hinder the collective benefits of progress.
• Solid ethical grounding to never lose sight, in a fast-changing context, of that sacred space for dignity and autonomy that we call privacy.