I doubt this explains everything, but it is a start.

https://www.orfonline.org/research/explainable-a-i-and-its-military-implications/

Managing Expectations: Explainable A.I. and its Military Implications

The potential for military use has often been the driving force of technological innovation around the world. In recent years, there has been a notable increase in the development and deployment of highly advanced disruptive technologies for defence purposes, and artificial intelligence (AI) has become the poster child for this trend. Only a few years ago, the current gamut of applications of AI in military operations would have been dismissed as fodder for fiction. Today, with advances in emerging technologies in the area of lethal autonomous weapons systems (LAWS) and the continuous integration of AI and machine learning (ML) into the back-end of existing military computing systems, military applications of AI systems around the world are only set to increase in number and intensity.[1] This surge is accompanied by new ideas of ensuring that the deployed military AI systems are more compatible with human use and have smaller margins of error. One such idea is the development of what is called eXplainable AI (XAI), i.e. AI and ML systems that make it possible for human users to understand, appropriately trust, and effectively manage AI.[2]

This brief explains why such systems are a necessity in the military, what XAI is and how it functions, examples of where and how it has been applied so far, and evaluates its use and regulation. The brief uses both primary and secondary research sources, including interviews with expert stakeholders from different geographies and disciplines, either currently or formerly from government, defence services, civil society, and academia. It aims to analyse the current status of XAI in the military, and pave the way for more targeted research.

“Look! I’ve still got the email you sent me ordering me to send $100,000 to that lawyer in Outbackistan!”

https://www.csoonline.com/article/3670548/why-business-email-compromise-still-tops-ransomware-for-total-losses.html#tk.rss_all

Why business email compromise still tops ransomware for total losses

While businesses are busy trying to protect themselves against ransomware attacks that spark headlines news, threat actors are sticking to one of the oldest and most effective hacking techniques—business email compromise (BEC).

Enterprise security has skewed toward ransomware in recent years, but FBI data highlights that  enterprises in aggregate are losing 51 times more money through BEC attacks. In 2021, BEC attacks in the US caused total losses of $2.4 billion, a 39% increase from 2020. In contrast, at the same time, companies in the US lost only $49.2 million to ransomware.

What guidelines or tests will be applied before software like this is used?

https://futurism.com/the-byte/professor-ai-predicts-crimes

PROFESSOR SAYS HE FORESEES NO ISSUES WITH HIS AI THAT PREDICTS CRIMES BEFORE THEY HAPPEN

A crime-prognosticating AI now exists, and it's been implemented in a number of American cities. But don't worry, it definitely won't be misused by city police forces — at least according to its lead creator, University of Chicago professor Ishanu Chattopadhyay.

Chattopadhyay recently sat down with BBC Science Focus to discuss the AI system, which, as a study published in the journal Human Behavior claims, can predict where and when a crime might occur with 80 to 90 percent accuracy. But whether the predictions hold up in the real world, while of course important, isn't necessarily the core question here: really, the question is whether AI can be successfully incorporated into a police force without abuses — and while Chattopadhyay believes his system can be, AI's track record in policing says otherwise.

There must be money in it…

https://www.theregister.com/2022/08/24/amazon_ring_surveillance/

Amazon has repackaged surveillance capitalism as reality TV

… A world of pervasive observation by invisible authorities with the power to judge, sentence and enforce their verdicts without appeal or even making themselves visible to the accused – maybe that's all just a source of endless comedy?

That's clearly what Amazon believes. It recently announced the launch of its MGM-branded series, "Ring Nation" – a light-hearted look at some of the funniest footage gathered by Amazon's vast network of spy eyes. According to Hollywood insider website Deadline, "The series will feature clips such as neighbors saving neighbors, marriage proposals, military reunions and silly animals."

This announcement – made just a few weeks after Amazon's admission of Ring footage being given to legal authorities without a warrant – simultaneously reads as both a failure to read the room, and the ultimate troll. "See what we've done here? We've turned Nineteen Eighty-Four's Room 101 into The Benny Hill Show!"

I wonder if we could tune it to give everyone an Australian accent? Irish? Japanese?

https://futurism.com/startup-ai-remove-call-center-accents

Cursed Startup Using AI to Remove Call Center Workers' Accents

When Boots Riley's "Sorry To Bother You" dropped in 2018, we knew it was only a matter of time before some of the film's dystopian predictions started to come true. But we've gotta admit, AI technology that converts realtime speech into a cookie cutter, white-coded accent is pushing it.

As SFGATE reports, a startup called Sanas is offering "accent translation" that makes call center employees — many of whom are hired overseas, where labor is cheaper, and hence have non-American accents — sound more palatable to American ears.

In case you are (also) curious…

https://www.bespacific.com/the-mar-a-lago-search-warrant-a-legal-introduction/

The Mar-a-Lago Search Warrant: A Legal Introduction

Follow up to previous posting, Trump Had More Than 300 Classified Documents at Mar-a-Lago, which included links to government documents pertaining to the FBI search and NARA’s repeated requests to secure the return of classified government documents, please see CRS Legal Sidebar The Mar-a-Lago Search Warrant: A Legal Introduction. August 23, 2022: “The Federal Bureau of Investigation (FBI) recently executed a search warrant at former President Donald Trump’s Mar-a-Lago property in Palm Beach, Florida. A magistrate judge of the U.S.District Court for the Southern District of Florida later unsealed the warrant at the Department of Justice’s request. which the former President did not oppose. The warrant authorized government officials to seize all “documents and records constituting evidence, contraband, fruits of crime, or other items illegally possessed in violation” of three federal statutes—18 U.S.C. §§ 793, 2071, and 1519. In addition to the warrant itself and its attachments, the court unsealed other material related to the search, including the cover sheet to the warrant application and an inventory of property seized. Proceedings are underway to unseal a redacted version of the affidavit supporting the warrant, and former President Trump filed a motion asking the court, among other things, to appoint a special master to oversee the government’s handling of the seized material. This Sidebar describes the process for and implications of obtaining a search warrant. It then examines the criminal offenses identified in the Mar-a-Lago warrant. Finally, this Sidebar analyzes presidential authority to declassify documents and the role of declassification for the crimes at issue…”