A snapshot of Russian disinformation, and its remaining audience. A look at ChatGPT.

At a glance.

• RT honcho says her broadcasts are getting out, no matter what.
• Russian propaganda and diminishing returns.
• An alternative account of the targeting at Makiivka.
• And an alternative account of Russian retaliation for Makiivaka.
• Perspective on ChatGPT.

RT honcho says her broadcasts are getting out, no matter what.

"They understand that it's important," RT chief Margarita Simonyan said on the Solovyov show, "We're not on Telegram on Western platforms. We exist in Russia, in Russian language, but our foreign language outlets are not. They are not on TikTok; that's over. We're now creeping down partisan roads. I won't divulge any other details, not to show our trump cards to their intel."

For what it's worth, we've experienced little to no difficulty in seeing what RT and the other Russian state outlets (TASS, Sputnik, etc.) are pushing, so we're not entirely sure what the blockage she's alluding to amounts to.

Russian propaganda and diminishing returns.

Russian propaganda seems increasingly inward-looking, directed toward keeping a domestic audience quiescent and docile in the face of increasing sacrifice, and even that effort may now be subject to the law of diminishing returns. The Atlantic Council assesses Moscow's disinformation efforts as, internationally speaking, largely a failure, and a failure that's compounded the war's larger reputational damage to Russia:

"At the same time, the damage to Russia’s reputation has already been done. Russia’s global standing has always relied heavily on international perceptions of the country as a major military power. This myth has now been ruthlessly exposed on the battlefields of Ukraine. Countries which had earlier felt obliged to remain on good terms with Russia now understandably feel they have little to fear, while those who previously saw Moscow as a powerful partner have been forced to rethink this relationship.

"Domestically, the consequences may be even more critical for Putin. Belief in Russia’s military strength served as the foundation stone of the country’s modern national identity. It was a source of patriotic pride that helped justify the often harsh living conditions and limited individual rights that all Russian citizens are forced to accept. This entire facade is now in danger of collapsing."

At this point the main lines of Russian disinformation are what an essay in Foreign Policy calls "memory diplomacy," that is, a propaganda line that trades upon a memory of the Second World War that gives Russia sole credit for victory over Nazi Germany, that whitewashes Stalin's complicity in opening European phases of that war, and that seeks to frame the Ukrainian state as the historical successor to the Nazis in a war that never really ended. The problem with this memory is that no one else shares it.

In the US, there's been opportunity for some reflective assessment of the effects Russian influence operations had on the 2016 Presidential election. That there was a significant attempt to move the election is beyond serious dispute. It now seems beyond dispute that those efforts showed very little payoff. Both the Washington Post ("minimal impact") and the Register ("weak sauce") have accounts of the investigation.

An alternative account of the targeting at Makiivka.

NEXTA, a video news outlet operated by Belarusian dissidents from Poland, has a video made by a Russian soldier who was wounded in the HIMARS strike against the temporary barracks in Makiivka. Contrary to Russian Ministry of Defense reports (which blamed soldiers' careless use of cellphones for betraying their position), the soldier (who subsequently died of his wounds) says that the troops had been assembled in the auditorium of the occupied school in order to hear President Putin's New Year's address. "All the locals," he said, knew the troops were being collected in the building. If accurate, the soldier's account places the onus of poor tactical practice and atrocious operations security squarely on the local commanders.

And an alternative account of Russian retaliation for Makiivaka.

Meanwhile, yesterday the Russian Ministry of Defense announced that it had killed over six-hundred Ukrainian soldiers in a "retaliatory" strike designed to exact retribution for Makiivka. The AP says "The Russian Defense Ministry said its missiles hit two temporary bases housing 1,300 Ukrainian troops in Kramatorsk, in the eastern Donetsk region, killing 600 of them. Ministry spokesman Igor Konashenkov said the strikes were retaliation for Ukraine’s attack in Makiivka, in which at least 89 Russian soldiers died. Serhii Cherevatyi, a spokesperson for Ukraine’s forces in the east, told The Associated Press that Russian strikes on Kramatorsk damaged only civilian infrastructure, adding: 'The armed forces of Ukraine weren’t affected.' Reports on the ground are consistent with Kyiv's account. They indicate that the strike was unsuccessful: the missiles failed to hit their targets, and in any case the Ukrainian soldiers weren't there. "On Sunday, reporters in Kramatorsk, Donbas, shared photos which showed the strikes missed their targets and further claimed no one was inside the barracks at the time of the blast," the Telegraph writes.

Perspective on ChatGPT.

ChatCPT, now out in an early version, has attracted a great deal of attention for its plausible imitation of human speech. Perhaps its most interesting feature is the easy natural language interaction it affords users, as Help Net Security notes. It can also be used to write code, although the results here seem less striking and impressive. Its ability to generate convincing phishing communications at scale is more impressive, as TechCrunch reports, and this capability for deception and impersonation is probably its most interesting feature over the short term. It's a natural language analogue of the deepfake technology that's been applied to images.

Boris Cipot, Senior Security Engineer at Synopsys Software Integrity Group, sent us an early assessment of the upsides and downsides of the kind of AI ChatGPT represents.

The security benefits.

“AI is already a part of Cybersecurity tools. What this now means and how much is just marketing is in the eye of the beholder, however there are active projects that are focused to detect and even mitigate possible threats with artificial intelligence.

“With the arrival of a technology like ChatGPT the use cases are vast. From helping with code checks, to lending a helping hand with writing scripts to pen testers and alike. Those are just simple generic use cases. There are for sure more specific tasks ChatGPT can fulfill in the detection and mitigation space.

“One point we must however understand is that even the part of ChatGPT that can produce malicious code can be used for good, for example pinpointing weak spots in software or systems to mitigate them. However, the more this AI will learn the more it will be able to then help the bad guys, and this is now the threat – the more you use it the more capable it becomes, the more it can help but also destroy. This is unfortunately a gap. Building a sense of moral into ChatGPT might help to prevent the misuse as ChatGPT would then decide whom it will help and work for or work with. But the question is then for whom the AI will decide.”

The security risks.

“Cybercrime, hacking, cracking, call it as you may, is a combination between knowledge and imagination. Knowledge delivers you the means to execute the attack, however imagination is the key element on how to do the attack. ChatGPT can be “used” in this regard from many angles: from writing texts that are of special interest to you and luring you into clicking links in the text that are malicious all to writing scripts, small programs that can be used for those malicious purposes.”