Experts warn that ISPs would use black-box AI to comply
Like our piece on Canada’s Cyber-Security Bill earlier this year, this item was flagged in my work for the Internet Society, Canada Chapter, but this post does not necessarily reflect the opinion of the ISCC.
It concerns a legal opinion drafted by the UK’s Open Rights Group (similar mission to Internet Society and other civil society groups), on that country’s planned “Online Harms Bill”.
Recall that here in Canada, we also have Online Harms legislation in the works. When Steven Guilbeault was Heritage Minister, he began crafting Bill C-36, which would seek (among other things), to designate political dissent as “hate speech”, criminalize the criticism of politicians and may require ISPs to implement “an internet kill switch to block websites deemed hurtful”.
The major red flag in the UK bill is the concept of “prior restraint“. In communications to the ISOC, Dr Monica Horten expressed Open Rights Group’s concerns that,
“This is the concept of banning content before publication, or blocking publication without due process, and may be interpreted as an ‘upload filter’.
Whilst the legal Advice does not directly address the provisions that affect encrypted messaging services, it could have implications. It addresses Clause 9 (2) which requires online providers to ‘prevent users encountering’ illegal content. The same language is used in Clause 111 which establishes the power for Ofcom to require encrypted messaging services to screen users’ communications. For this reason, we would like to draw your attention to it.
The Advice warns that the Online Safety Bill could fundamentally alter online communication by enforcing ‘prior restraint through AI, stifling freedom of expression by blocking content deemed ‘illegal’ without – explanation, notification or due process‘ for the censored user.
The Advice considers that Clause 9 (2) provides for state-sanctioned ‘prior restraint’ on freedom of expression through the use of proprietorial processes by private companies. This will lead to interference with freedom of expression which is not “prescribed by law” – meaning that it would not be sufficiently accessible, clear and foreseeable to allow individuals to regulate their conduct accordingly for the purposes of Article 10 of the European Convention on Human Rights.
As Thomas Sowell would say, “Oh dear, where to begin?”
We saw under the Covid years how pre-emptive shaping of permissible speech on the part of governments, the Big Tech platforms and various private “fact checking” rackets, resulted in wholesale censorship, deplatforming, and even coordinated defamation and professional harm.
In short, people had their lives and careers destroyed for speaking out, for questioning the “official” narrative, or for proposing alternative mitigation methodologies than those espoused by unelected, unaccountable technocrats.
In many cases, those prevailing, official narratives have turned out to be fallacious or incorrect to the point of being somewhere on a spectrum between misinformation and institutionalized gaslighting.
Governments, Big Tech, and the corporate media should be slinking off into a corner and licking the wounds of their now shattered credibility, while trying to figure out ways to rebuild trust amongst a public that they have all betrayed beyond redemption.
Instead, they are trying to ram through new legislation, more rules, and opaque algos that would seek to further penalize dissent, criminalize non-conforming thought, and as Dr. Horten spells out plainly: violate universal human rights.
It would also thrust the implementation phase of this Brave New World onto ISPs, service providers and communications services, which for the most part, are built to do the exact opposite of what these new initiatives propose. In many cases we are now dealing with decentralized, open source protocols for which such new legislation would be pointless and unenforcible, if not absurd.
All so that they can double-down on their own failures and escape facing any accountability for getting every meaningful talking point around the pandemic dead wrong.
Meanwhile, In Canada…
Moving the Online Safety legislation down the field is now under the purview of Pablo Rodriguez, Canada’s current former Heritage Minister (he’s out in a cabinet reshuffle moments ago), who was currently embroiled in the debacle of having single-handedly deplatformed the Canadian corporate media from Big Tech platforms with the deeply flawed Bill C-18 (C-36’s original architect Steven Guilbeault is currently Environment Minister).
On it’s own C-36 would be bad enough, but in concert with the myriad other pieces of legislation
- C-11 put the Internet under the regulatory oversight of the CRTC
- C-26 (the Cyber-Security bill) would grant sweeping, unsupervised power to bureaucrats and even quasi-government agencies to literally command and censor internet communications by decree
- C-18 – aforementioned – trying to strong arm tech platforms into subsidizing Canada’s flailing corporate media
…we continue to see some very disturbing convergences both here at home and internationally.
Subscribe to the easyDNS #AxisOfEasy newsletter for a weekly digest and podcast covering internet security, censorship and privacy, and follow @AxisOfEasy on Twitter, or Nostr via npub1za32zx2xg99nfut4ch5y7km8vpumvsj48d4hakjnslhvjuzq2v6qm5qkml
Leave a Reply