• AWS hit in latest Amazon job cuts
• Redundancies are also looming at Telecom Italia
• Liberty Global wants near total ownership of Belgium’s Telenet

In today’s industry news roundup: Cloud services giant AWS is to suffer a headcount reduction in Amazon’s latest round of job cuts; the knives are also out at Telecom Italia, it seems; Liberty Global wants to buy out other shareholders in Belgian operator Telenet; and more. 

Following in the footsteps of Meta, Amazon has become the second tech giant to announce a second round of job cuts, having already axed 18,000 jobs in January, and this time staff from its highly successful Amazon Web Services (AWS) cloud division are impacted. The company’s CEO, Andy Jassy, sent out a staff memo earlier today announcing plans to cut a further 9,000 positions in the next few weeks. The layoffs will mainly affect the AWS team, its People Experience and Technology Solutions (PXT Solutions) team, the company’s advertising unit and its Twitch streaming service team. “This was a difficult decision, but one that we think is best for the company long term… [given] the uncertain economy in which we reside and the uncertainty that exists in the near future,” wrote Jassy. He added that following several years of hiring “a significant amount” of people when it “made sense given what was happening in our businesses and the economy as a whole… the overriding tenet of our annual planning this year was to be leaner while doing so in a way that enables us to still invest robustly in the key long-term customer experiences that we believe can meaningfully improve customers’ lives and Amazon as a whole.” Amazon will confirm which specific roles will be affected by mid-to-late April and Jassy insisted the company will support those who lose their jobs with packages, including a separation payment, transitional health insurance benefits and external job placement support. “I remain very optimistic about the future and the myriad of opportunities we have, both in our largest businesses, stores and AWS, and our newer customer experiences and businesses in which we’re investing”, he added. Google parent Alphabet, Microsoft, Twitter, Snap and Ericsson have also significantly reduced their headcounts recently.

Telecom Italia (TIM) is in discussions with unions about a 5% workforce reduction, which would affect 2,000 jobs, through the introduction of a voluntary early-retirement scheme, Reuters has reported. TIM is current seeking any way possible to reduce its costs and cut its debt pile of €25bn, and is currently engaged in a process that, it hopes, will result in the sale of NetCo, its fixed line and international networks business unit – see Crunch time for Telecom Italia as NetCo bids fall short.

Liberty Global plans to increase its stake in Belgian cable and mobile service provider Telenet to more than 96% by offering minority shareholders €22 in cash per share, a premium of 59% compared with Telenet’s closing price on 15 March 2023, and a premium of 52% compared to the volume-weighted average trading price of Telenet’s shares during the previous month. Liberty Global currently holds a controlling 59.18% stake in Telenet and wants to own the rest of the shares, apart from the 3.12% stake held by Telenet itself. The board of the Belgian operator “unanimously supports and recommends the intended offer,” it noted in a statement issued on Tuesday. Telenet’s share price gained almost 40% to reach €21.24 on the Euronext exchange following the announcement. For the full year 2022, Telenet reported a 3% year-on-year increase in revenues to almost €2.7bn and a 4% dip in operating profit to €576m. It ended last year with 1.73 million broadband, 1.7 million video/TV and 2.94 million mobile customers, while almost 824,000 customers take both fixed and mobile services, a 10% year-on-year increase. News of Liberty Global’s move came a day after the European Commission approved Orange Belgium’s acquisition of Belgian cable operator Voo – see Orange boosts its convergence play with green light for Belgian acquisition.

The government of Hungary is set to increase its control over the national telecom sector, as state-owned Corvinus International Investment has acquired a minority stake in PPF Telecom Group’s operator Yettel Hungary through a share-swap deal. As part of the agreement, Corvinus International Investment is exchanging a 19.5% stake in Vodafone Hungary for a 25% stake in PPF Telecom Group’s subsidiaries in Hungary (owned by 4iG Group), which include telecom operator Yettel Hungary, wholesale provider of active and passive telecoms infrastructure CETIN Hungary, and property company Yettel Real Estate Hungary. Corvinus International Investment’s stake in Vodafone Hungary has been acquired by Antenna Hungária for HUF17.7bn (US$48m), which has boosted Antenna Hungária’s direct controlling stake in the Hungarian operator to 70.5%, while the state-owned company holds the remaining 29.5%. The Hungarian government has claimed that the move will diversify its investment portfolio in the sector while allowing the state to “more forcefully and efficiently contribute to the development of the industry,” reported Reuters.

In the workers’ paradise that is North Korea, mobile telephony, such as it is, controlled and policed as it is, and expensive as it is, has been slow to spread amongst the ordinary public. Those devices and services that could be bought (handsets and SIMs alike) were priced in US dollars because the elite of the regime needed (and still need) hard currency from any source possible. And, of course, those that bought them were duly registered as their owners. Now, as the number of handsets in circulation slowly increases, pop-up stalls have, well, popped-up in markets across North Korea, selling pre-loved handsets that remain registered in the names of the original owners. Between 2017 and 2020, the North Korean authorities permitted the release of two or three new so-called “smartphone” models a year. As new ones were bought, individuals sold on their old handsets. Many of these so-called “daepo” phones were not re-registered and so became, to some extent, untraceable. The word ‘daepo’ has five meanings in the Korean language, the most common being “cannon” as in weaponry or “grog” as in booze (in general rum, and water in particular). Now it seems the sheer number of daepo phones in circulation is becoming a problem in a country where absolute control over everyone is the number one priority. So now the national telecoms law, that already prescribes and proscribes so much, is to be extended to make it illegal for anyone to own a daepo. The story comes from Daily NK, an online newspaper based in Seoul in South Korea that covers various aspects of life in North Korea (with information sourced from a network of brave and very necessarily anonymous North Korean informants). The person who broke the above story across the borders and into the wider world wrote, “The law was revised because phones used under borrowed names have spread across the entire country due to the indiscriminate sale of mobile phones and SIM cards. There continue to be cases where people can’t be punished because the phones they are using illegally aren’t in their own names.” So now it is illegal for the phones to be repaired or serviced, and service centres are compelled to report any person coming in for work to be done on a mobile phone. The regime rails daily about young people destroying their linguistic purity by using South Korean idioms or slang, and police and other agencies routinely wipe services and apps from any handsets that are deemed to be capable of getting round North Korea’s home-grown mobile security systems (in other words, all of them that don’t meet with official approval for any reason). Those found guilty of violating the new law, and in North Korea the conviction rate in courts is 99.999% recurring for any offence in any jurisdiction, get seven years hard labour in a “re-education camp”. Harsh.

Pan-African network operator and digital service provider Liquid Intelligent Technologies has acquired Cairo, Egypt-based cloud and security services specialist Cysiv MEA (formerly SecureMisr) for an undisclosed sum. Cysiv MEA will adopt the Liquid C2 brand name, which is used for the Liquid group’s cloud operations, “to align it with its global cloud and cybersecurity identity,” noted Liquid Intelligent Technologies, which plans to “significantly grow the Egyptian business by tapping into the wealth of local tech talent, making Egypt a key hub for the Middle East and North Africa (MENA) region.” David Behr, CEO of Liquid C2, added: “Liquid recognises the critical role Cysiv MEA has been playing in the cloud and cyber security industry in Egypt and the region. Our main task as a group is to support them in bringing more cyber security tools for our customers as they face an increasingly hostile global threat environment from cybercriminals and nation-state sponsored attackers. This will ensure that their business is protected whilst also meeting the demands for global compliance requirements.” Read more. 

The latest Internet Satisfaction Report survey, commissioned by Amdocs and carried out by Dynata, a research and analytics company headquartered in Connecticut, US, has found that 89% of Americans say they now have reliable broadband internet access, notes Telecompetitor. What’s more, they regard broadband access as much of a necessity of life as utility services such as gas, water and electricity. Meanwhile, demand for, and reliance on, broadband connectivity has accelerated to the extent that the number of households with more than nine internet connected devices has doubled in less than two years (though actual numbers were not specified in the report). Unsurprisingly, the number of devices per home correlates closely with household income, so the Amdocs survey shows that while 41% of families with annual incomes in excess of US$150,000 actually have more than nine internet connected devices, just 13% of “low-income” homes (defined in the report as earning less than $50,000 per annum) have fewer than nine devices. The survey, which canvassed the views of 1,000 US citizens, also found that, despite occasional outages and other connection issues, 49% of families are happy with their home internet. What’s more, 40% reported having no issues or difficulties whatsoever with their broadband, while a mere 11% said they had either occasional or more frequent connectivity problems. Interestingly, and as usual, politics plays its part in attitudes to broadband availability. Thus, 61% of respondents said it is important that the persistent digital divide, which still bedevils some regions and conurbations of the US, finally be bridged before the 2024 presidential election. Meanwhile, 26%, presumably in light of the furore about claims that the 2020 election was “stolen” from the loser Donald Trump, said it is "extremely important" to do so, while 35% felt it is of some importance, 26% were “unsure” and 4% didn’t care either way. Turning to other areas, 61% believe that AI, virtual reality, augmented reality and the metaverse will actually increase the digital divide rather than help bridge it. The remaining 39% were not concerned. Amusingly, there was no unanimity whatsoever on who should be responsible for closing the gap. Anthony Goonetilleke, Amdocs’ group president of technology and head of strategy, commented: “Failing to ensure universal connectivity only serves to exacerbate the digital divide, leaving those on the lower end of the socioeconomic ladder at an even greater disadvantage in their personal, professional, and academic pursuits. Given the role that technology and connectivity play in driving the global economy, it’s our collective responsibility to remain persistent in our efforts to close this gap.” As for the usual bones of contention – broadband speed and reliability – 52% of respondents said they would be willing to pay more for the former, while 38% would pay extra for the latter. The report also hinted that consumer attitudes to ISPs may have mellowed somewhat and approval of their services improved subsequent to the privations caused by the lockdowns that characterised the early months of the Covid-19 pandemic. During that period the reputations of many US ISPs took a battering, often with good reason.

Less than six months after its launch, streaming giant Netflix has reportedly hit one million monthly active users (MOUs) for its Basic with Adverts subscription service in the US. According to internal data obtained by Bloomberg, the milestone was achieved two months after the service’s launch in November 2022 and the user base rose by more than 500% in its first month, in a sign of the times as the cost of living crisis takes hold. The report also claimed that most of the subscribers opting for the cheaper, ad-supported plan, are either new to Netflix or are relapsing customers rather than users who have downgraded from more pricey plans. In the US, the basic plan with ads starts at $6.99 per month compared with the entry-level price of $9.99 for the advert-free service.

When it comes to stopping AI ripping-off artists’ works, the only way isn't ethics. Academics are helping artists to counter the dismal and pervasive trend for modern day thieves to use generative AI to rip-off original works, feed the stolen data sets into tools that have been built to read and copy visual styles, and then quickly and cheaply churn-out ersatz simulacra for ill-gotten gains. The Glaze project, underway at the University of Chicago in the US, has just launched a free, non-commercial app that artists can download to help prevent the theft of “artistic IP” via a new “cloaking” technique. The beta version of the new app is now available and works by adding miniscule “perturbations” to paintings and other artworks that will make the AI model misread the data that breaks down the essence of an individual artist’s particular style. The result is a generated impersonation that is nothing like anything the original artist would have produced, and therefore less valuable. TechCrunch, the US online publication that covers high-tech and start-up companies, spoke with Ben Zhao, professor of computer science at the University of Chicago and the project leader. He told TechCrunch, “What we do is try to understand how the AI model perceives its own version of what artistic style is. And then we basically work in that dimension – to distort what the model sees as a particular style. So it’s not so much that there’s a hidden message or blocking of anything… It is, basically, learning how to speak the language of the machine learning model, and using its own language — distorting what it sees of the art images in such a way that it actually has a minimal impact on how humans see. This comes from a fundamental gap between how AI perceives the world and how we perceive the world.” The root of the cheap “plausible pastiche” problem is that the makers of generative AI models rely on easy access to the public internet for data to train their models. Thus, artists who have displayed their work online which, usually, is about the only thing that can inexpensively do to attract an audience and potential buyers of their work, are regarded as fair game. Zhao added, “We’re trying to address this issue of artists feeling like they cannot share their art online. Particularly independent artists who are no longer able to post, promote and advertise their own work for commission – and that’s really their livelihood. So just the fact they can feel like they’re safer, and that it becomes much harder for someone to mimic them, means that we’ve really accomplished our goal. The large majority of artists out there can use this, they can feel much better about how they promote their own work and they can continue on with their careers and avoid most of the impact of the threat of AI models mimicking their style.”

- The staff, TelecomTV