I wrote this week on the latest skirmishing in the crypto wars. End-to-end encryption has conquered the world. Governments continue to push back in defence of lawful intercept. But the debate around technical solutions has changed little in years. economist.com/international/…
A few sources for this piece. In 2018 & 2021 a pair of GCHQ officials proposed various solutions for reconciling gov't interception with end-to-end encryption. The first piece proposed that govts could be added secretly to particular communications: lawfaremedia.org/article/princi…
The second proposed various types of "client-side scanning"—automatically comparing content on a device to a stored library of illegal material—as a way of detecting child sexual abuse material on social media sites without intercepting content in transit. arxiv.org/abs/2207.09506
Cryptographers argued that the first idea—adding gov't to chats as a secret participant, or a 'ghost protocol'—had significant downsides, because it'd require pushing code changes to all users. blog.cryptographyengineering.com/2018/12/17/on-…
In a paper in 2021, many prominent cryptographers also pushed back against client-side scanning, arguing that it would introduce vulnerabilities, that it could be dangerously expanded in scope or exploited by attackers. arxiv.org/abs/2110.07450
.@AlecMuffett has a good (long) primer on end-to-end encryption and why it's so hard to devise means to lawfully access content that does not in effect undermine encryption as a whole. alecmuffett.com/alecm/e2e-prim…
This paper looks at the encryption debate in India, which has demanded that message services identify the "originator" of a message. WhatsApp has said that would undermine encryption and has threatened to withdraw from India. carnegieindia.org/research/2023/…
In 2021 @ciaranmartinoxf gave a brilliant speech on these - increasingly vitriolic - debates around lawful access v encryption. "Some experts say, in effect, that the government is arguing not against a policy decision, but against mathematics." bsg.ox.ac.uk/sites/default/…
Governments are still asking tech companies to limit or modify the use of encryption. In April 32 European police chiefs said: "end-to-end encryption is being rolled out in a way that undermines their ability to investigate crime and keep the public safe." nationalcrimeagency.gov.uk/news/european-…
In April also a coalition of 15 law enforcement agencies including Interpol, FBI, NCA and others said much the same. They asked tech companies to "make tangible steps towards possible solutions that we know exist." nationalcrimeagency.gov.uk/statement-on-e…
The European Union continues to explore a very ambitious client-side scanning scheme ("Chat Control"), though there is no consensus on it yet. A note on this was presented to ambassadors as recently as yesterday. euractiv.com/section/law-en…
Matthew Green, a cryptographer, laid out a critique of the EU scheme below.
A CNAS tabletop exercise. "This study finds that a hypothetical, protracted U.S.-PRC conflict creates conditions under which nonstrategic nuclear weapons use is both appealing to the PRC and difficult to manage for the United States" cnas.org/publications/r…
"once nuclear escalation in the Indo-Pacific occurs, reciprocal tactical nuclear exchanges may continue, but not necessarily lead to general nuclear war." cnas.org/publications/r…
"These findings reflect the fundamental differences of deterrence in the emerging Indo-Pacific era, where distinct geography, targets, and capabilities make limited nuclear escalation potentially more tolerable than in the Cold War era." cnas.org/publications/r…
“Russia has trained its navy to target sites deep inside Europe with nuclear-capable missiles in a potential conflict with Nato, according to secret files…Maps of targets as far-flung as the west coast of France and Barrow-in-Furness” on.ft.com/4fzYtM2
‘The document notes the navy’s “high manoeuvrability” allows it to conduct “sudden & pre-emptive blows” and “massive missile strikes . . . from various directions”. It adds that nuclear weapons are “as a rule” designated for use “in combination with other means of destruction”..’
‘The presentation also references the option of a so-called demonstration strike — detonating a nuclear weapon in a remote area “in a period of immediate threat of aggression” before an actual conflict to scare western countries.’ on.ft.com/4fzYtM2
Kommersant reports from Kursk. General sense being chaos & abandonment by Moscow. “I want to understand where our state is at all? Where is the administration? I wish they could talk to us. We don't know anything at all… there's no one.” kommersant.ru/doc/6890223
“Why wasn't there a fucking evacuation? Everyone ran away as best they could - under fire, under kamikaze drones! Why did they lie on TV to the last? They said that the situation was stable” kommersant.ru/doc/6890223?fr…
“Tell the state that we want to see them. Let the state tell us the truth - what should we expect? Will we return to our homes or can we say goodbye to them? Well, at least some crumb of honest information from the state!” kommersant.ru/doc/6890223?fr…
Struck by the fact that we can barely work out with confidence which Ukrainian brigades & what proportion of them are properly in in Kursk, let alone if anything is in reserve & what precisely is happening. OSINT can be absolutely magical, and sometimes incredibly limiting.
And so beware wild evidence free speculation however enjoyable it feels.
This seems to have been taken as a complaint. It is not. It is a comparison between the relatively high levels of battlefield transparency we have become used to, esp. when front lines are static, to when they become fluid & when Ukr prioritises surprise.
“The son of immigrants.” For these people, we’ll never really be British.
A couple of other lines in that which stand out. Goodwin cites: “The creeping sense of lawlessness.”
Reality: “Crime in England and Wales has fallen to its lowest level on record…Incidents of violent crime have dropped by 28% in the year to June 2023.” bbc.co.uk/news/uk-671619…
“ONS reported that victimisation rates shown by the CSEW have been decreasing in the long term. They peaked in the year ending December 1995, when 4.7% of adults were a victim of violent crime. Rates have remained below 2% since the year ending March 2014” lordslibrary.parliament.uk/trends-in-viol…
🧵 While I am going through notes, there is also a good discussion here between @StaciePettyjohn & @NarangVipin, the US' Secretary of Defense for Space Policy (and an MIT professor). Some interesting bits on nuclear policy/strategy incl. NATO nuclear posture.
Narang says scenario "flipped" from cold war, where NATO planned nuclear first use. "NATO is the conventionally superior [one]. And so the muscle memory we have from the Cold War is not actually not as applicable...the central challenge is how do we deter Ru first employment "
Narang: "we are [now] developing that muscle memory, how do we deter Russian first employment, particularly non-strategic nuclear weapons, or I like to refer to them ... as treaty unaccountable nuclear weapons". Says NATO has made major progress in two years on this.