Staying in touch with a loved one after their death is the promise of several start-ups using the powers of artificial intelligence, though not without raising ethical questions.

Ryu Sun-yun sits in front of a microphone and a giant screen, where her husband, who died a few months earlier, appears.

“Sweetheart, it’s me,” the man on the screen tells her in a video demo. In tears, she answers him, and a semblance of conversation begins.

When Lee Byeong-hwal learned he had terminal cancer, the 76-year-old South Korean asked startup DeepBrain AI to create a digital replica using several hours of video.

“We don’t create new content” such as sentences that the deceased would have never uttered or at least written and validated during their lifetime, said Joseph Murphy, head of development at DeepBrain AI, about the “Rememory” program.

“I’ll call it a niche part of our business. It’s not a growth area for us,” he cautioned.

The idea is the same for StoryFile, a company that uses 92-year-old “Star Trek” actor William Shatner to market its site.

“Our approach is to capture the wonder of an individual, then use the AI tools,” said Stephen Smith, boss of StoryFile, which claims several thousand users of its Life service.

Entrepreneur Pratik Desai caused a stir a few months ago when he suggested people save audio or video of “your parents, elders and loved ones,” estimating that by “the end of this year” it would be possible to create an autonomous avatar of a deceased person, and that he was working on a project to this end.

The message posted on Twitter set off a storm, to the point that, a few days later, he denied being “a ghoul.”

“This is a very personal topic and I sincerely apologize for hurting people,” he said.

“It’s a very fine ethical area that we’re taking with great care,” Smith said.

After the death of her best friend in a car accident in 2015, Russian engineer Eugenia Kyuda, who emigrated to California, created a “chatbot” named Roman like her dead friend, which was fed with thousands of text messages he had sent to loved ones.

Two years later Kyuda launched Replika, which offers personalized conversational robots, among the most sophisticated on the market.

But despite the Roman precedent, Replika “is not a platform made to recreate a lost loved one,” a spokesperson, said.

Somnium Space, based in London, wants to create virtual clones while users are still alive so that they then can exist in a parallel universe after their death.

“It’s not for everyone,” CEO Artur Sychov conceded in a video posted on YouTube about his product, Live Forever, which he is announcing for the end of the year.

“Do I want to meet my grandfather who’s in AI? I don’t know. But those who want that will be able to,” he added.

Thanks to generative AI, the technology is there to allow avatars of departed loved ones to say things they never said when they were alive.

“I think these are philosophical challenges, not technical challenges,” said Murphy of DeepBrainAI.

“I would say that is a line right now that we do not plan on crossing, but who knows what the future holds?” he added.

“I think it can be helpful to interact with an AI version of a person in order to get closure — particularly in situations where grief was complicated by abuse or trauma,” Candi Cann, a professor at Baylor University who is currently researching this topic in South Korea.

Mari Dias, a professor of medical psychology at Johnson & Wales University, has asked many of her bereaved patients about virtual contact with their loved ones.

“The most common answer is ‘I don’t trust AI. I’m afraid it’s going to say something I’m not going to accept.’ … I get the impression that they think they don’t have control” over what the avatar does.
…

Resisting the hype, Apple defied most predictions this week and made no mention of artificial intelligence when it unveiled its latest slate of new products, including its Vision Pro mixed reality headset.

Generative AI has become the tech world’s biggest buzzword since Microsoft-backed OpenAI released ChatGPT late last year, revealing the capabilities of the emerging technology. 

ChatGPT opened the world’s eyes to the idea that computers can churn out complex, human-level content using simple prompts, giving amateurs the talents of tech geeks, artists or speechwriters. 

Apple has laid low as Microsoft and Google raced out announcements on how generative AI will revolutionize its products, from online search to word processing and retouching images.

During the recent earnings season, tech CEOs peppered mentions of AI into their every phrase, eager to reassure investors that they wouldn’t miss Silicon Valley’s next big chapter.

Apple has chosen to be much more discreet and, in its closely watched keynote address to the World Developers conference in California, never once mentioned AI specifically.

“Apple ghosts the generative AI revolution,” said a headline in Wired Magazine after the event. 

‘Not necessarily AI?’

Arguments vary on why Apple has chosen a more subtle approach. 

For one, Apple follows other critics who have long been wary of the catchall “AI” term believing that it is too vague and unhelpfully evokes dystopian nightmares of killer robots and human subjugation to machines. 

For this reason, some companies – including TikTok or Facebook’s Meta – roll out AI innovations, but without necessarily touting them as such. 

“We do integrate it into our products [but] people don’t necessarily think about it as AI,” Apple CEO Tim Cook told ABC News this week.

Indeed, AI was actually very much part of Apple’s annual jamboree on Monday, but it required a level of technical know-how to notice.

In one instance, Apple’s head of software said “on-device machine learning” would enhance autocorrect for iPhone messaging when he could have just as well said AI.

Apple’s autocorrect innovation drew giggles with the promise of iPhones no longer correcting common expletives.

“In those moments where you just want to type a ‘ducking’ word, well, the keyboard will learn it, too,” said Craig Federighi.

Autocorrect will also learn from your writing style, helping it guide suggestions, using AI technology similar to what powers ChatGPT.

In another example, a new iPhone app called Journal, an interactive diary, would use “on-device machine learning … to inspire your writing,” Apple said, again not referring to AI when other companies would have.

But AI will also play a major role in the Vision Pro headset when it is released next year, helping, for example, generate a user’s digital persona for video-conferencing.

‘Not much effort’

For some analysts, the non-mention of AI is an acknowledgement by Apple that it lost ground against rivals. 

“They haven’t put much effort into it,” independent tech analyst Rob Enderle told AFP. 

“I think they just kind of felt that AI was off into the future and it wasn’t anything surprising,” he added. 

The glitchy performance of Apple’s chatbot Siri, which was launched a decade ago, has also fed the feeling that the smartphone giant doesn’t get AI. 

“I think most people would agree that Apple lost its edge with Siri. That’s probably the most obvious way they fell behind,” said Insider Intelligence principal analyst Yory Wurmser. 

But Wurmser also insisted that Apple is primarily a device company and that AI, which is software, will always be “the means rather than the ends for a great user experience” on its premium devices.

In this vein, for analyst Dan Ives of Wedbush Securities, the release of Apple’s Vision Pro headset was in itself an AI play, even if it wasn’t explicitly spelled out that way.

“We continue to strongly believe this is the first step in a broader strategy for Apple to build out a generative AI driven app ecosystem” on the Vision Pro, he said. 
…

There are renewed concerns North Korea’s army of hackers is targeting financial institutions to prop up the regime in Pyongyang and possibly fund its weapons programs.

A report published Tuesday by the cybersecurity firm Recorded Future finds North Korean aligned actors have been spoofing well-known financial firms in Japan, Vietnam and the United States, sending out emails and documents that, if opened, could grant the hackers access to critical systems.

“The targeting of investment banking and venture capital firms may expose sensitive or confidential information of these entities or their customers,” according to the report by Recorded Future’s Insikt Group.

“[It] may result in legal or regulatory action, jeopardize pending business negotiations or agreements, or expose information damaging to the company’s strategic investment portfolio,” it said.

The report said the most recent cluster of activity took place between September 2022 and March 2023, making use of three new internet addresses and two old addresses, and more than 20 domain names.

Some of the domains imitated those used by the targeted financial institutions.

Recorded Future’s named the group behind the attacks Threat Activity Group 71 (TAG-71), which is also known as APT38, Bluenoroff, Stardust Chollima and the Lazarus Group.

This past April, the U.S. sanctioned three individuals associated with the Lazarus Group, accusing them of helping North Korea launder stolen virtual currencies and turn it into cash.

U.S. Treasury officials levied additional sanctions just last month against North Korea’s Technical Reconnaissance Bureau, which develops tools and operations to be carried out by the Lazarus Group.

The Lazarus Group is believed to be responsible for the largest theft of virtual currency to date, stealing approximately $620 million connected to a popular online game in Match 2022.

Earlier this month, U.S. and South Korean agencies issued a warning about another set of North Korean cyber actors impersonating think tanks, academic institutions and journalists in an ongoing attempt to collect intelligence.

 
…

Japan announced Tuesday that it joined the United States and Australia in signing a $95 million undersea cable project that will connect East Micronesia island nations to improve networks in the Indo-Pacific region where China is increasingly expanding its influence.

The approximately 2,250-kilometer (1,400-mile) undersea cable will connect the state of Kosrae in the Federated State of Micronesia, Tarawa in Kiribati and Nauru to the existing cable landing point located in Pohnpei in Micronesia, according to the Japanese Foreign Ministry.

Japan, the United States and Australia have stepped up cooperation with the Pacific Islands, apparently to counter efforts by Beijing to expand its security and economic influence in the region.

In a joint statement, the parties said next steps involve a final survey and design and manufacturing of the cable, whose width is about that of a garden hose. The completion is expected around 2025.

The announcement comes just over two weeks after leaders of the Quad, a security alliance of Japan, the United States, Australia and India, emphasized the importance of undersea cables as a critical component of communications infrastructure and the foundation for internet connectivity.

“Secure and resilient digital connectivity has never been more important,” Matthew Murray, a senior official in the U.S. State Department’s Bureau of East Asian and Pacific Affairs, said in a statement. “The United States is delighted to be part of this project bringing our region closer together.”

NEC Corp., which won the contract after a competitive tender, said the cable will ensure high-speed, high-quality and more secure communications for residents, businesses and governments in the region, while contributing to improved digital connectivity and economic development.

The cable will connect more than 100,000 people across the three Pacific countries, according to Kazuya Endo, director general of the international cooperation bureau at the Japanese Foreign Ministry.

 
…

The U.S. Department of Homeland Security is looking at how artificial intelligence can be used to help identify goods made with child or forced labor and prevent those goods from entering the country. VOA’s Julie Taboh has more. VOA footage by Adam Greenbaum.
…

Top Chinese officials told Elon Musk about plans to launch new regulations on artificial intelligence on his recent trip to the Asian giant, the tech billionaire said Monday, in his first comments on the two-day visit.

The Twitter owner and Tesla CEO — one of the world’s richest men — held meetings with senior officials in Beijing and employees in Shanghai last week.

“Something that is worth noting is that on my recent trip to China, with the senior leadership there, we had, I think, some very productive discussions on artificial intelligence risks, and the need for some oversight or regulation,” Musk said. “And my understanding from those conversations is that China will be initiating AI regulation in China.”

Praised China

Musk, whose extensive interests in China have long raised eyebrows in Washington, spoke about the exchange in a livestreamed Twitter discussion with Democratic presidential hopeful and vaccine conspiracy theorist Robert Kennedy Jr., the nephew of the late U.S. President John F. Kennedy.

Musk did not tweet while in China and Tesla has not released readouts of Musk’s meeting with officials.

But official Chinese channels said he lavished praise on the country, including for its “vitality and promise,” and expressed “full confidence in the China market.”

Several Chinese companies have been rushing to develop AI services that can mimic human speech since San Francisco-based OpenAI launched ChatGPT in November.

But rapid advancements have stoked global alarm over the technology’s potential for disinformation and misuse.

Musk didn’t elaborate on his discussions in China but was likely referring to a sweeping draft law requiring new AI products to undergo a security assessment before release and a process ensuring that they reflect “core socialist values.”

The “Administrative Measures for Generative Artificial Intelligence Services” edict bans content promoting “terrorist or extremist propaganda,” “ethnic hatred” or “other content that may disrupt economic and social order.”

Under Beijing’s highly centralized political system, the measures are almost certain to become law.

Describes meetings as ‘promising’

Musk has caused controversy by suggesting the self-ruled island of Taiwan should become part of China — a stance that was welcomed by Chinese officials but which deeply angered Taipei.

The 51-year-old South African native described his meetings in China as “very promising.”

“I pointed out that if there is a digital super intelligence that is overwhelmingly powerful, developed in China, it is actually a risk to the sovereignty of the Chinese government,” he said. “And I think they took that concern to heart.”
…

The European Union is pushing online platforms like Google and Meta to step up the fight against false information by adding labels to text, photos and other content generated by artificial intelligence, a top official said Monday.

EU Commission Vice President Vera Jourova said the ability of a new generation of AI chatbots to create complex content and visuals in seconds raises “fresh challenges for the fight against disinformation.”

Jourova said she asked Google, Meta, Microsoft, TikTok and other tech companies that have signed up to the 27-nation bloc’s voluntary agreement on combating disinformation to dedicate efforts to tackling the AI problem.

Online platforms that have integrated generative AI into their services, such as Microsoft’s Bing search engine and Google’s Bard chatbot, should build safeguards to prevent “malicious actors” from generating disinformation, Jourova said at a briefing in Brussels.

Companies offering services that have the potential to spread AI-generated disinformation should roll out technology to “recognize such content and clearly label this to users,” she said.

Jourova said EU regulations are aimed at protecting free speech, but when it comes to AI, “I don’t see any right for the machines to have the freedom of speech.”

The swift rise of generative AI technology, which has the capability to produce human-like text, images and video, has amazed many and alarmed others with its potential to transform many aspects of daily life. Europe has taken a lead role in the global movement to regulate artificial intelligence with its AI Act, but the legislation still needs final approval and won’t take effect for several years.

Officials in the EU, which is bringing in a separate set of rules this year to safeguard people from harmful online content, are worried that they need to act faster to keep up with the rapid development of generative artificial intelligence.

The voluntary commitments in the disinformation code will soon become legal obligations under the EU’s Digital Services Act, which will force the biggest tech companies by the end of August to better police their platforms to protect users from hate speech, disinformation and other harmful material.

Jourova said, however, that those companies should start labeling AI-generated content immediately.

Most of those digital giants are already signed up to the EU code, which requires companies to measure their work on combating disinformation and issue regular reports on their progress.

Twitter dropped out last month in what appeared to be the latest move by Elon Musk to loosen restrictions at the social media company after he bought it last year.

The exit drew a stern rebuke, with Jourova calling it a mistake.

“Twitter has chosen the hard way. They chose confrontation,” she said. “Make no mistake, by leaving the code, Twitter has attracted a lot of attention and its actions and compliance with EU law will be scrutinized vigorously and urgently.”
…

In collaboration with the Ukrainian government, the U.S. Agency for International Development, or USAID, has created an app that connects Ukrainians with their government so they can access public services — and use of the app’s code has expanded to different countries. Iryna Matviichuk has the story, narrated by Anna Rice.
…

Amazon agreed Wednesday to pay a $25 million civil penalty to settle Federal Trade Commission allegations it violated a child privacy law and deceived parents by keeping for years kids’ voice and location data recorded by its popular Alexa voice assistant.

Separately, the company agreed to pay $5.8 million in customer refunds for alleged privacy violations involving its doorbell camera Ring.

The Alexa-related action orders Amazon to overhaul its data deletion practices and impose stricter, more transparent privacy measures. It also obliges the tech giant to delete certain data collected by its internet-connected digital assistant, which people use for everything from checking the weather to playing games and queueing up music.

“Amazon’s history of misleading parents, keeping children’s recordings indefinitely, and flouting parents’ deletion requests violated COPPA (the Child Online Privacy Protection Act) and sacrificed privacy for profits,” Samuel Levine, the FCT consumer protection chief, said in a statement. The 1998 law is designed to shield children from online harms.

FTC Commissioner Alvaro Bedoya said in a statement that “when parents asked Amazon to delete their kids’ Alexa voice data, the company did not delete all of it.”

The agency ordered the company to delete inactive child accounts as well as certain voice and geolocation data.

Amazon kept the kids’ data to refine its voice recognition algorithm, the artificial intelligence behind Alexa, which powers Echo and other smart speakers, Bedoya said. The FTC complaint sends a message to all tech companies who are “sprinting to do the same” amid fierce competition in developing AI datasets, he added.

“Nothing is more visceral to a parent than the sound of their child’s voice,” tweeted Bedoya, the father of two small children.

Amazon said last month that it has sold more than a half-billion Alexa-enabled devices globally and that use of the service increased 35% last year.

In the Ring case, the FTC says Amazon’s home security camera subsidiary let employees and contractors access consumers’ private videos and provided lax security practices that enabled hackers to take control of some accounts.

Amazon bought California-based Ring in 2018, and many of the violations alleged by the FTC predate the acquisition. Under the FTC’s order, Ring is required to pay $5.8 million that would be used for consumer refunds.

Amazon said it disagreed with the FTC’s claims on both Alexa and Ring and denied violating the law. But it said the settlements “put these matters behind us.”

“Our devices and services are built to protect customers’ privacy, and to provide customers with control over their experience,” the Seattle-based company said.

In addition to the fine in the Alexa case, the proposed order prohibits Amazon from using deleted geolocation and voice information to create or improve any data product. The order also requires Amazon to create a privacy program for its use of geolocation information.

The proposed orders must be approved by federal judges.

FTC commissioners had unanimously voted to file the charges against Amazon in both cases.
…

The International Energy Agency says Chinese car manufacturers are emerging as a major force in the global electric car market, with more than 50% of all electric cars on roads worldwide now produced in China. Spain is the second-largest vehicle manufacturer in Europe after Germany and its market has become a target for Chinese automakers. From Barcelona, Alfonso Beato has this report, narrated by Marcus Harton.
…

SpaceX’s Starlink, the satellite communications service started by billionaire Elon Musk, now has a Defense Department contract to buy those satellite services for Ukraine, the Pentagon said Thursday.  

“We continue to work with a range of global partners to ensure Ukraine has the resilient satellite and communication capabilities they need. Satellite communications constitute a vital layer in Ukraine’s overall communications network and the department contracts with Starlink for services of this type,” the Pentagon said in a statement.

Starlink has been used by Ukrainian troops for a variety of efforts, including battlefield communications.  

SpaceX, through private donations and under a separate contract with a U.S. foreign aid agency, has been providing Ukrainians and the country’s military with Starlink internet service, a fast-growing network of more than 4,000 satellites in low Earth orbit, since the beginning of the war in 2022.

The Pentagon contract is a boon for SpaceX after Musk, the company’s CEO, said in October it could not afford to indefinitely fund Starlink in Ukraine, an effort he said cost $20 million a month to maintain.

Russia has tried to cut off and jam internet services in Ukraine, including attempts to block Starlink in the region, though SpaceX has countered those attacks by hardening the service’s software.

The Pentagon did not disclose the terms of the contract, which Bloomberg reported earlier on Thursday, “for operational security reasons and due to the critical nature of these systems.”
…

The semiconductor trade war between Washington and Beijing may ensnare Seoul as South Korea must decide between backing its closest ally or embracing a lucrative export opportunity presented by China, its top trading partner. 

The decision will reveal how closely South Korea is aligned with the U.S., its second-largest export market, experts said. 

The dilemma facing Seoul emerged after China announced that it was banning the use of U.S.-based Micron Technology’s broad range of computer memory and storage technologies. 

Liu Pengyu, a Chinese Embassy spokesperson in Washington, told VOA’s Korean Service on May 24 that Beijing’s cybersecurity regulators had assessed that Micron’s chips “pose a major security risk to China’s key information infrastructure supply chain and impact China’s national security.” 

The ban echoed that set by the U.S. on China’s Huawei Technologies in May 2019, when the Trump administration cited security concerns related to the company’s wireless networking equipment, especially those related to 5G. The Biden administration in November 2022 banned approvals of new telecommunications equipment from Huawei and ZTE because the products pose “an unacceptable risk” to U.S. national security.

U.S. Representative Mike Gallagher, the Republican chairman of the House Select Committee on Strategic Competition between the United States and the Chinese Communist Party, has called for South Korea to “act to prevent backfilling” the market gap left by Micron.

Litmus test

The U.S. has been trying to block China’s access to the technology needed to make advanced chips that can be used to modernize its military. Micron’s chips are used by Chinese industries that assemble consumer electronics such as smartphones. Although Beijing is funding the development of home-grown advanced chips such as those used in artificial intelligence applications, China’s chipmakers, for now, manufacture simpler products such as those used in home appliances.

Seoul’s decision on whether to dissuade its top chipmakers such as Samsung or SK Hynix from selling chips to China could indicate how closely South Korean President Yoon Suk Yeol is aligned with Washington. 

“This would certainly be a litmus test to see if Seoul and other allies are willing to support Washington’s policies designed to slow China’s technology growth,” said Andrew Yeo, the SK-Korea Foundation chair in Korea Studies at Brookings Institution.

Robert Rapson, who served as charge d’affaires and deputy chief of mission at the U.S. embassy in Seoul, 2018-2021, said, “This is the first real test of the Yoon administration’s policy of enhanced alignment with the U.S. on China.” 

He continued, “In other words, will [South] Korea sacrifice core economic, commercial interests of its flagship high-tech companies in keeping with [Washington’s] policy and U.S. wishes?”

He added that Seoul has the right to seek “some credit or offset” from Washington if it blocks backfilling the Micron gap. 

A business decision

A spokesperson for the South Korean Foreign Ministry told VOA’s Korean Service on Tuesday that the government “plans to continue efforts to protect the interest of our companies through cooperation with relevant agencies and engagements with diplomatic missions abroad.”  

South Korea sent 55% of its semiconductor exports to China last year even as  

its semiconductor exports have been in a steep decline since August 2022, according to a Bank of Korea report released on Tuesday, cited by Business Korea.

Robert Manning, a senior fellow at the Stimson Center’s Reimagining U.S. Grand Strategy Project, said “As the security environment in Northeast Asia has become fraught with North Korea’s provocative nuclear efforts and Chinese economic coercion, the U.S.-ROK alliance has become more vital to Seoul.” South Korea’s official name is the Republic of Korea (ROK).

“South Korea will [need to] sacrifice to a degree to sustain broad alignment with the U.S.,” Manning said. “But South Korea has its own interests so there are likely to be limits.” 

Troy Stangarone, senior director at Korea Economic Institute, said, “While China might face short-term shortage in chips if Samsung and SK Hynix withheld capacity, the ultimate result would only be the further expansion of domestic Chinese semiconductor firms which undermine U.S. long-term goals and potentially the very firms the United States is working with to improve its own supply chains.”

Dennis Wilder, senior director for East Asia affairs at the White House’s National Security Council during the George W. Bush administration, said, “This is a business decision, and it really should, in my view, be left to the South Korean companies to make this business decision.”  

Wilder continued, “But it’s far more important for South Korea to align with the United States on the very high-end semiconductor chips and the attempts to keep things out of the hands of the Chinese military that can help modernize.”  

Beijing’s ban came on the last day of the Group of Seven countries summit on May 19-21. The group agreed to de-risk the global economy and diversify trade away from China in an effort to counter its economic coercion. This is defined as “a threatened or actual imposition of economic costs by a state on a target with the objective of extracting a policy concession,” according to testimony by Bonnie Glaser, managing director, of the German Marshall Fund Indo-Pacific program, before the Congressional-Executive Commission on China.

U.S. Commerce Secretary Gina Raimondo said on Saturday that Washington “firmly opposes” China’s ban on Micron. She made the remark at a press conference held after the meeting of the U.S.-led Indo-Pacific Economic Framework (IPEF) that China sees as a body aimed at countering its economic rise. 

On Monday, an article in Chinese state-run media Global Times said it would be “natural” for South Korea’s chipmakers to export to fill the market void left by the Micron ban.

“There is no possibility for South Korea to replace its chips with other goods in its exports to China,” the report said. 

And on Sunday, Bloomberg quoted an unidentified source familiar with the situation as saying South Korea will veer away from supplying chips to China.

South Korea’s exports to China in April were $9.52 billion while exports to the U.S. reached $9.18 billion, according to the Trade Ministry’s latest data. The gap between South Korea’s exports to China and the U.S. narrowed to just $340 million in April from $1.15 billion in January driven by a strong dollar and EV demand.
…