Glass Panels Exterior of the Microsoft Building

From October, Microsoft will check everything you do on your Windows computer

From 30 September, new terms and conditions will apply to Microsoft customers. If you do not adhere to a vague code of conduct when using products of the quasi-monopolist, your account can be blocked and you can lose access to all paid or free services and your data stored there. Recourse is available on a discretionary basis through Microsoft.

According to Microsoft’s new “services agreement” the following applies:

“Severe or repeated violations of our policies, including violations of our child exploitation and abuse imagery policy, may result in account suspension. Sometimes a suspension may be permanent. Under permanent suspension, the owner of the suspended profile forfeits all licenses, subscriptions, membership time, and Microsoft account balances. “

You will be allowed to appeal against this to Microsoft. Many of the examples given of prohibited behaviour are criminal acts, from child pornography to phishing. But there are also terms in the Code of Conduct that we have learned are extremely elastic, such as “hate speech” and “offensive”. It has become customary to call it hate speech, if someone criticises the government or a person in a harsh way. It is also relatively easy to fall foul of copyright protection without being criminally inclined.

Microsoft promises to only impose proportionate penalties for serious and repeated violations. But I haven’t found anything in the set of rules that would offer a halfway credible guarantee against arbitrary sanctios of critics of governments and their narratives by social media platforms. Under pressure from governments they  have over-censored and over-blocked in a completely non-transparent manner, sometimes for the smallest or nonexisting offences.

Invitation to governments

The platforms have even set up special pages for governments, through which these can transmit their censorship requests in a privileged manner. This is likely to happen with Microsoft under this services agreement

This is highly convenient for governments: They no longer need to prove criminal misconduct against a recalcitrant artist or author if they want to silence him or her. It is enough to point out to Microsoft that one of the rubber clauses of the “services agreement” has allegedly been violated. This possibility is even explicitly addressed on Microsofts webpage on enforcement:

“We leverage reports from users, governments, and trusted flaggers to bring potential policy violations to our attention.”

Clearly, a company must respond when a government agency points out continued criminal behaviour using its products. But why should a company specifically invite governments and authorities to alert it to violations of its own policies that go far beyond that?

That is an open invitation to abuse power. Why would Microsoft mess with a government that abuses its privileged right to suggest censorship, why insist on a different assessment in individual cases? A government, especially a powerful one, can cause a lot more trouble for the company than a customer with an unfairly blocked account.

As a recent example, the social media platform Rumble responded harshly to the British Parliament, from which it had received a letter asking if and when it would demonetise the content of Russell Brand, the entertainer who has been an influencial critic of government policies. Brand is facing allegations of sexual misconduct. Rumble reminded parliament of the presumption of innocence until conviction as a central pillar of the rule of law. (Demonetising means that no advertising can be placed.) Tiktok and other platform operators also received the letter from the British parliament.

Vague rules open door to arbitrariness

Even generating nude images is a breach of the Microsoft Services Agreement. Nude painting, nude photography, lovers posing for each other, all this can lead to account suspension. According to the wording, it is enough to just generate or store “offensive” language or images, you don’t even have to spread it.

I have not found any commitment by Microsoft to explain and justify sanctions. So the company can act as has been common practice for some time with social media platforms and financial service providers like Paypal. A user account is simply blocked with a reference to unspecified rule violations and the customer is left to guess why and how best to object. One is reminded of Kafka’s Josef K., who is persecuted and sanctioned by the authorities without being charged, and whose attempts to defend himself end nowhere, as he cannot find his prosecutors or the judges.

Automated total surveillance

On the page on the “enforcement processes” one learns:

“At Microsoft, we use a combination of automated technology and trained human reviewers to find and take action against any content or conduct that violates our terms and policies. (…)  We also use machine-learning technologies like text-based classifiers, image classifiers, and grooming detection techniques to discover content or conduct shared through Microsoft hosted consumer services that may be illegal or break our policies. Lastly, we leverage reports from users, governments, and trusted flaggers to bring potential policy violations to our attention.”

Recently, public prosecutors in Germany have taken action against artists and authors for allegedly glorifying Nazism, even though they had been quoting fascist and Nazis speech and symbols in an obviously critical manner. If public prosecutors can be so dense, do we really want to trust an artificial intelligence and some underpaid moderators in India that they will correctly classify whether someone has created some supposedly hateful content so spread hate or in a critical or satirical way.


As Microsoft constantly sifts through all the user data stored in its cloud in an automated way, users are exposed to a sophisticated, almost all-encompassing surveillance infrastructure that is constantly evolving based on its vast data. It will become more and more complex and opaque in the process. Users have no way of checking what the algorithms are looking for, and what patterns they are assembling and storing. For the intelligence services and authorities, such a surveillance infrastructure is a sitting duck.

That’s why I personally have been using a Linux operating system for a long time and – as far as I know – no Microsoft products. If I were more familiar with computers, I would provide some guidance. But with a little asking around, most of you should be able to find someone who can help you say goodbye to Microsoft.

Originally published by Money and more

Suggest a correction

Similar Posts