There is a deep underlying connection between the three news stories covered during yesterday’s episode of The Communications Risk Show: the loss of trust in the messages we read and hear, and the difficulty of recovering that trust.
- Considerable fuss is being made about Tata Communications trying to buy a business which owns The Campaign Registry, the anti-spam registry for SMS and MMS that is used by the biggest mobile operators in the USA. National registries are emerging as a popular response to spam and scams sent as bulk A2P messages. The principle is simple: registries are very large allow lists. If your message and your business has not been vetted and registered in advance, then the telcos connected to the registry will not deliver the message to its end users. The idea is appealing, but who oversees the people who run the registries, to ensure they do their job properly? And is there a national security risk in allowing a foreign company to run a registry that could influence which political robotexts are seen during the run-up to an election?
- Anybody who has used social media this year is likely to be aware of the excitable claims made on behalf of ChatGPT and other AI-powered language models. The supposed upside is that anybody can provide these models with a few ill-considered ungrammatical notes and the machine will turn them into sparkling prose that Ernest Hemingway would envy. But any short cut delivered by technology innovation is just as likely to be exploited by cheats and liars as anyone else. How are you supposed to trust what you are reading on the web if there is an overwhelming surge in output from machines that are designed to mimic real people? And how do we avoid a plagiarism armageddon as crooks and idle students avoid detection when they copy somebody else’s work? A machine that can swap synonyms, reorder words, and merge content from several sources to disguise the fact that no human has actually written anything new is a gift to intellectual property thieves and fraudsters who want many variations of the same essential scam.
- Nobody likes receiving robocalls that hide their true origin. But then, nobody wants to go to a lot of trouble to validate the origin of calls that somebody else will receive, especially if the favor is not returned for calls that come the other way. International voice carriers are caught in the crossfire as different countries propose and implement incompatible methods of validating calls. Philippe Millet of i3forum (pictured, right) was our expert guest, telling us about the reasons why telcos, regulators and associations focused on trust in the telecoms ecosystem need to get involved in the One Consortium, the new global nonprofit entity he has set up to harmonize the methods that will be applied to call validation. The first online meeting of the One Consortium will occur on September 7, and Philippe told us about who has already committed to take part.
The words you hear through your phone’s speaker, the words you read on your phone’s screen, and the words you read on a computer screen are all morphing into output from apps that run over the internet. They will all be subject to the same race between the technology to increase the spread of words and the technology to suppress the spread of words. Sometimes the suppression will be necessary, and other times it will be excessive. Sometimes crooks will get their messages out, and sometimes authoritarians will stop us receiving messages we needed to hear. The severity of these issues will rapidly escalate in the near future. The way societies, governments and businesses respond to these challenges will define the extent of the freedom and safety enjoyed by ourselves and our children. That is why we talk about these topics.
We want you to be involved in the conversation too. Our shows are streamed live every Wednesday at 4pm UK, making it easy to save time in your diary. But if you miss a show you can always find the recording at the dedicated website, tv.commsrisk.com. You can also replay yesterday’s show below.
