Prysmian Paid $1.15 Billion for Channell Commercial

South Korea is tightening preparations for the nine-hour Suneung college entrance exam, taken by more than 500,000 students. The push follows a cheating case in which tax-accountant exam top scorer Han Won-seok said he photographed questions with a miniature camera, sent them to generative AI and checked answers on his smartphone. Police are investigating, and the Human Resources Development Service of Korea apologized and plans to file a criminal complaint. Officials will place specially trained supervisors at each of 1,312 test centers and inspect students’ glasses, including for cameras, screens, charging ports and power buttons on models such as Meta and Rokid.

Authorities are also weighing collection of smart devices, suspending public Wi-Fi, detecting Bluetooth and limiting mobile networks. But signal shutdowns could disrupt nearby residents and emergency calls, and experts have raised legal concerns. Education Minister Choi Kyo-jin rejected postponing the exam, while officials say they lack time and legal authority to screen every student with metal detectors before the 2027 test. The government plans to announce further anti-cheating measures and is examining protections for other state-run exams.

Elsewhere in AI infrastructure, Nokia, Nscale, Skeleton, Telia and Verda have formed AI Gigafactory Northern, a proposed Nordic-Baltic network combining AI computing, sovereign cloud, connectivity and lower-carbon energy. The consortium plans to bid through a European high-performance-computing program; research partners VTT and KTH will contribute expertise. The initiative comes as the EU seeks to mobilize €30 billion for AI investment. In Africa, a report says inadequate data-centre capacity, limited investment and high energy costs are holding back AI adoption, and calls for funding and expertise to expand modern facilities.

At COP31 on Nov. 16, 2026, Univers announced partnerships with the World Green Building Council, the Mærsk Mc-Kinney Møller Center for Zero Carbon Shipping and XIBER Energy Solutions. Its Physical AI platform will be applied to buildings, maritime operations and industry, and the company introduced a framework for assessing AI’s value in physical operations. Separately, Prysmian CEO Massimo Battaini says the cable maker will revisit a possible U.S. dual listing in the coming months. North America now generates 40% of its revenue, versus 20% in 2024; its regional growth includes the 2025 purchase of Channell Commercial for $1.15 billion. Data-centre spending is lifting demand for power and fibre-optic cables, and Battaini says he sees no evidence of an AI bubble.

President Donald Trump’s Asia trip is expected to put nuclear arms discussions and trade in focus, including talks with China’s Xi Jinping and possible U.S.-China-Russia arms negotiations. Before traveling to China, Trump is expected to sign a Thailand trade deal and announce investments in aviation, defense and technology. Nvidia is also in focus: the FBI is investigating allegations involving the advanced-chip maker, while lawmakers call for limits on China’s access to its chips.

Other reports cover AI’s effects on work and research. An economic FAQ argues that existing measures—low unemployment, workers’ organizing and bargaining rights, a stronger safety net and progressive taxation, particularly on capital income—may better protect workers than AI-specific rules alone. Researchers introduced ModHarness-RL, which lets training harnesses evolve through learned Compression and Steering programs; tests on CodeSearch, ALFWorld and BIRD SQL with 1.5B–8B models improved held-out performance over fixed harnesses in most settings. AI forecasting systems can approach skilled humans over short horizons, but long-term accuracy and benchmark reliability remain uncertain, as does public trust in opaque predictions.

Two additional developments underline practical risks and market interest. Ransomnews found that some old Bitly links on X had been repurposed to redirect visitors to unrelated sites; Bitly said bad actors exploited certain inactive links and that it suspended the accounts involved, though many recreated links still worked when reported. MarketBeat’s AI-stock screener named Hut 8, Tempus AI, Cyabra, SentinelOne, Fermi, InterDigital and Appian as stocks to watch.

Key Takeaways

  • South Korea will assign specially trained supervisors to all 1,312 Suneung test centers and inspect glasses after an AI-assisted cheating case involving a hidden camera and smartphone.
  • Officials are considering device collection, Wi-Fi or mobile-network restrictions and Bluetooth detection, but signal blocks raise public-safety and legal concerns.
  • Han Won-seok, who ranked first in both rounds of a national tax-accountant exam, is under police investigation; the exam authority plans a criminal complaint.
  • Nokia, Nscale, Skeleton, Telia and Verda formed AI Gigafactory Northern to propose connected AI infrastructure across the Nordic and Baltic region.
  • A report says Africa needs investment in modern data centres to address capacity, infrastructure and energy-cost constraints on AI adoption.
  • Univers announced COP31 partnerships spanning buildings, shipping and industry, alongside a framework to assess AI’s value in physical operations.
  • Prysmian plans to revisit a possible U.S. dual listing in the coming months; North America provides 40% of revenue, and its 2025 Channell Commercial acquisition cost $1.15 billion.
  • Trump’s Asia trip puts nuclear discussions and trade in focus; Nvidia is also in focus amid an FBI investigation and calls in Congress to restrict China’s access to its chips.
  • ModHarness-RL improved held-out results over fixed training harnesses in most reported settings; AI forecasting remains less certain over the long term.
  • Reports also flag reused old Bitly links on X, potential AI-related worker disruption, and seven AI stocks named by MarketBeat: Hut 8, Tempus AI, Cyabra, SentinelOne, Fermi, InterDigital and Appian.

South Korea considers stronger checks to stop AI cheating

South Korea is weighing stronger security measures after miniature cameras and AI smart glasses appeared in cheating schemes. Experts recommend screening for electronic devices, changing entry procedures, adding trained proctors and creating legal authority and funding for these steps. A Ministry of Education official said metal detectors could not be used to search every student before the 2027 College Scholastic Ability Test because there is not enough time or legal authority. The government is also discussing blocking wireless signals at test sites, while officials plan to announce new anti-cheating measures on the 17th.

AI cheating scandal raises alarm before South Korea’s college exam

South Korea is preparing for its College Scholastic Ability Test, or Suneung, which more than 500,000 students will take on Thursday. Concern grew after 34-year-old Han Won-seok admitted using a hidden camera, a smartphone and generative AI to cheat on a national tax accountant exam. Education Minister Choi Kyo-jin rejected calls to delay the college exam, saying a postponement would cause greater disruption. Officials plan to place specially trained supervisors at each of 1,300 test sites and are considering collecting smart devices and shutting down public Wi-Fi.

South Korea trains staff to spot AI glasses before college exam

South Korea is stepping up checks for AI-powered glasses ahead of Thursday’s College Scholastic Ability Test, known as Suneung. At least one specially trained official will work at each of the country’s 1,312 test centers to help identify the devices. Proctors will look for features such as cameras, thick frames, screens, charging ports and power buttons, including on Meta and Rokid models. Prime Minister Han Seong-sook also directed officials to collect students’ smart devices and examine whether public Wi-Fi could be disabled during the exam. The government plans to investigate cheating reports after the test as well.

South Korea considers new steps to stop AI cheating on college exam

South Korean officials are considering stronger security for Thursday’s College Scholastic Ability Test, or Suneung, which more than half a million students will take. The concern increased after Han Won-seok, the top scorer on a national tax accountant exam, confessed to using a hidden camera, smartphone and generative AI to get answers. Education Minister Choi Kyo-jin rejected calls to postpone the exam, while officials are considering collecting smart devices and shutting down public Wi-Fi at test sites. The government is also discussing limits on telecommunications, but signal jamming could affect nearby areas and emergency calls. Police are investigating Han’s case.

AI cheating case prompts urgent exam security plans in South Korea

South Korea is preparing for its nine-hour Suneung college entrance exam on Thursday, which more than half a million students will take. A man who ranked first on two national tax accountant tests admitted using a miniature camera and smartphone to get answers with AI. Officials are considering thermal cameras, suspending public Wi-Fi and limiting mobile signals, but experts warn that blocking signals could disrupt nearby residents and emergency calls. Education Minister Choi Kyo-jin rejected calls to reschedule the exam. The case has also raised questions about possible AI cheating in other tests, including law school admissions and university transfer exams.

South Korea weighs communication blocks to stop AI cheating on college exam

South Korea is considering ways to block communications at the College Scholastic Ability Test (CSAT) as concerns grow about cheating with AI and smart devices. The Ministry of Education and the Ministry of Science and ICT discussed cutting off Wi-Fi, 5G and LTE, pausing test takers’ phone lines, and detecting Bluetooth signals. Education officials said it is difficult to prevent all cheating without blocking communications, but telecom experts warned that network shutdowns could disrupt service for the public or break current laws. The government is also reviewing AI cheating prevention for other state-run exams, including authorized test sites and smart checkpoints.

South Korea steps up efforts to prevent AI cheating on college entrance exam

South Korean officials are concerned that AI and smart devices could help students cheat on the national college entrance exam, known as the CSAT. The exam plays a major role in deciding which universities students can attend and is a source of stress for families. The government says it is using AI-detecting software and security staff to monitor the test. Officials remain worried that students may find ways around these measures.

South Korea plans to inspect glasses at college entrance exam

South Korea plans to inspect all glasses worn by students taking its national college entrance exam this week. The move followed an unnamed student’s confession that they used AI to cheat on the 2022 university entrance exam. The confession raised concerns about exam fairness and the use of technology to cheat. A Ministry of Education spokesperson called the inspections necessary to protect the exam’s fairness.

South Korean police investigate top tax exam scorer over AI cheating

South Korean police are investigating a candidate who admitted to using AI to cheat in the national tax accountant exam. The candidate ranked first in both rounds, held in April and July, and scored 397.5 out of 400 in the first round. He said he used a miniature camera to photograph questions, sent them to AI, and checked answers on his smartphone. The Human Resources Development Service of Korea apologized and plans to file a criminal complaint, while the presidential office said officials would investigate suspected AI cheating across national qualification exams. The government has also increased efforts to catch people using smart glasses.

European partners plan Nordic and Baltic AI infrastructure network

Nokia, Nscale, Skeleton, Telia and Verda have formed AI Gigafactory Northern, a consortium to develop connected AI infrastructure across the Nordic and Baltic region. The planned network will combine AI computing, sovereign cloud services, high-capacity connectivity and decarbonized energy for customers across Europe. The partners intend to submit a bid under a program published by the European High-Performance Computing Joint Undertaking. Research partners VTT Technical Research Centre of Finland and KTH Royal Institute of Technology will contribute expertise in areas including AI, quantum computing and skills development. The announcement says the project is being developed as the EU seeks to mobilize €30 billion for AI investment through InvestAI.

Univers expands Physical AI partnerships across key industries

Univers announced three partnerships at COP31 on Nov. 16, 2026, with the World Green Building Council, the Mærsk Mc-Kinney Møller Center for Zero Carbon Shipping and XIBER Energy Solutions. Singapore’s Minister for Sustainability and the Environment Grace Fu witnessed the signing ceremony. The work will apply Univers’s Physical AI platform to buildings, maritime operations and industry, helping organizations connect operational technology, information technology and AI to improve complex systems. Univers also launched its Return on Intelligence framework, which aims to help businesses assess the value AI creates in physical operations.

Reused Bitly links redirect old posts to unexpected sites

Bitly links in old posts on X were reused to send visitors to websites unrelated to the original posts. Ransomnews found links from Burger King and McDonald’s tweets redirected to Wikipedia pages, while some links from 2009 to 2013 Trump tweets led to other destinations. Bitly said an issue allowed bad actors to repurpose certain old, inactive links and that it suspended the accounts involved. The report said the activity also affected links in posts by Bill Gates, Justin Trudeau, Mike Pence, the FBI and the Defense Department. Some links were deactivated, but many recreated links still worked at the time of the report.

How AI could affect jobs and the wider economy

This FAQ examines how artificial intelligence could affect U.S. workers, labor markets and the broader economy. It says concerns about job losses and falling wages deserve attention, but argues that policies designed specifically for AI may not keep pace with how businesses adopt the tools. The article points to existing policy approaches, including keeping unemployment low, protecting workers’ rights to organize and bargain, strengthening the social safety net and applying progressive taxes, especially to income from capital. It also notes that AI used in a labor market where many workers already face insecurity could add more disruption unless policies improve workers’ position.

ModHarness-RL lets AI training harnesses evolve with models

Researchers introduced ModHarness-RL, a method that treats the harness around an agentic reinforcement learning model as a program that can be learned and updated during training. Its harness records interactions as typed blocks and uses Compression programs to build the model’s context and Steering programs to guide behavior when it goes off course. The system proposes programs from recent training examples, tests them by running them, and checks them against past examples. Tests on CodeSearch, ALFWorld and BIRD SQL used models from 1.5B to 8B. Compression and Steering improved held-out performance over fixed harnesses in most settings.

Africa needs more data centres to support AI growth

A report by Stephen Nkrumah, dated Nov. 16, 2026, says Africa’s data centres lack the capacity and infrastructure to keep up with growing demand for AI services. Limited investment, inadequate facilities and high energy costs contribute to a gap between supply and demand, and the report says this is one reason AI adoption in Africa trails other regions. It calls for investment in modern, efficient data centres, with governments, businesses and international organizations providing funding and expertise. The report says expansion could support greater AI adoption, economic growth and competitiveness.

Trump’s Asia trip puts nuclear talks and Nvidia scrutiny in focus

President Donald Trump’s Asia trip comes as the White House seeks diplomatic and economic wins after the midterm elections. He is expected to discuss reducing nuclear weapons, which could lead to a new arms agreement among the United States, China and Russia. The wars in Iran and Ukraine may also shape talks with Chinese President Xi Jinping. Before traveling to China, Trump is expected to sign a trade deal with Thailand and announce investments in aviation, defense and technology. The trip also comes as the FBI investigates allegations involving Nvidia, a leading maker of advanced AI chips, and lawmakers call for limits on China’s access to its chips.

AI data center demand prompts Prysmian to revisit US listing

Prysmian, a cable maker listed in Milan, plans to restart its review of a possible US dual listing in the coming months, CEO Massimo Battaini said Monday. North America now brings in 40% of the company’s revenue, up from 20% in 2024. Its growth in the region includes the 2025 purchase of California-based Channell Commercial for $1.15 billion. Battaini said rising data center spending is lifting demand for power and fiber-optic cables, and he has seen “no evidence of an AI bubble.” He noted that market conditions remain uncertain.

Seven AI stocks to watch, according to MarketBeat

MarketBeat’s stock screener named seven artificial intelligence stocks to watch: Hut 8, Tempus AI, Cyabra, SentinelOne, Fermi, InterDigital and Appian. The article describes AI stocks as shares in companies substantially involved in developing products or services related to artificial intelligence.

AI forecasting systems approach top human forecasters, but gaps remain

Artificial intelligence forecasting systems can roughly match skilled human forecasters over the short term, but their long-term accuracy remains uncertain. AI can review large amounts of information quickly, though one researcher said current performance depends heavily on the quality of web searches. Experts warn that tests can give misleading results when systems have access to information about outcomes or when benchmarks are not carefully designed. Researchers also say current AI forecasts are broadly reasonable, but not ingenious, and that comparisons with human forecasters may fall within a margin of error. Trust remains a concern because people may not understand how the systems reach their predictions.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

AI-assisted cheating Suneung exam security Generative AI Smart glasses AI infrastructure AI Gigafactory Northern Sovereign cloud Data centers AI adoption in Africa Physical AI AI and energy AI chips Nvidia AI and employment ModHarness-RL AI forecasting Cybersecurity AI stocks Prysmian AI regulation

Comments

Loading...