An unusual pair of news anchors in Venezuela has emerged—El Pana and La Chama. These AI-generated figures, designed to look and sound realistic, are the creation of Connectas, a Colombia-based organisation. The ‘Operation Retweet’ initiative aims to disseminate news from several independent Venezuelan media outlets while protecting journalists from government repression. The project’s director, Carlos Huertas, explained that using AI allows them to bypass the escalating risks real reporters face in the country.
Why does it matter?
The Venezuelan government has been cracking down on journalists, protesters, and opposition figures amid a disputed election, with at least ten journalists arrested since mid-June, eight of whom remain imprisoned on severe charges. The crackdown is part of a broader effort to stifle dissent in response to the ongoing election dispute between President Nicolas Maduro and opposition candidate Edmundo Gonzalez. While Maduro, who has been in power since 2013, claims victory with the backing of the Supreme Court and electoral authority, the opposition argues that their candidate won by a large margin.
Protests over the election have resulted in 27 deaths and over 2,400 arrests. The situation has drawn international concern, with many questioning the election’s fairness and calling for the release of full vote tallies. Despite the government’s efforts to suppress dissent, the AI news anchors symbolise creative resistance, delivering news without putting human reporters at further risk.
California’s efforts to regulate the use of digital replicas of performers took a significant step forward with the passage of AB 1836 in the state Senate. The new bill mandates that studios obtain explicit consent from the estates of deceased performers before creating digital replicas for use in films, TV shows, video games, and other media. The move comes just days after the California legislature passed AB 2602, which enforces similar consent requirements for living actors.
SAG-AFTRA Statement on Today's Passing of CA Assembly Bill 1836: "For those who would use the digital replicas of deceased performers in films, TV shows, videogames, audiobooks, sound recordings and more, without first getting the consent of those performers’ estates, … 1/3
SAG-AFTRA, the union representing film and television performers, has strongly advocated for these measures, emphasising the importance of protecting performers’ rights in the digital age. In a statement released after the Senate’s approval of AB 1836, the union described the bill as a ‘legislative priority’ and urged Governor Gavin Newsom to sign it into law. The union’s stance highlights the growing concern over the unauthorised use of digital replicas, particularly as technology makes it increasingly easy to recreate performers’ likenesses long after they have passed away, keeping the audience concerned and aware of the issue.
If signed into law, AB 1836 would ensure that the estates of deceased performers have control over how their likenesses are used, potentially setting a precedent for other states to follow. However, the bill also raises practical challenges, such as determining who has the authority to grant consent on behalf of the deceased, which could complicate its implementation. The bill reflects a broader push within the entertainment industry to establish clear legal protections against exploiting living and deceased performers in the rapidly evolving digital landscape.
Alongside the AI bill, the passing of bill AB 1836 underscores California’s role as a leader in entertainment industry legislation, particularly in areas where technology intersects with performers’ rights. As the debate over digital replicas continues, the potential impact of AB 1836 on the industry could have far-reaching implications, keeping the audience engaged and interested in the future of entertainment law.
South Korean police have initiated an investigation into the messaging platform Telegram, probing its potential involvement in the spread of sexually explicit deepfake content. Authorities in South Korea are focusing on whether the platform may have facilitated the distribution of illegal material, following concerns raised by media reports.
Telegram responded by stating that it moderates harmful content on its platform through AI tools, proactive monitoring, and user reports. The company said millions of pieces of harmful content are removed each day.
South Korean authorities are determined to address the issue, with heightened efforts underway to fight the rising number of deepfake crimes, which have caused widespread alarm across the country.
OpenAI recently unveiled the Model Spec, a comprehensive framework designed to guide the behaviour of its GPT models in the OpenAI API and ChatGPT. The document is a crucial resource for researchers and data labellers involved in reinforcement learning from human feedback (RLHF), ensuring that models align with user intent and adhere to ethical standards.
The Model Spec is organised into three main components: Objectives provide broad directional goals, Rules establish specific instructions to prevent harmful outcomes and maintain legality, and Defaults offer basic style guidance and allow user flexibility while ensuring consistency.
The initiative serves multiple important purposes. It provides a framework for businesses to implement ethical AI, improve customer service quality, navigate regulations, and gain a competitive advantage through reliable AI systems. The Spec also addresses common issues by preventing users from prompting the model to ignore instructions and providing guidance on how models should refuse tasks.
OpenAI’s Model Spec represents a significant advancement in AI models’ fine-tuning and ethical alignment. As a living document, it will evolve based on community feedback and practical applications, contributing to the broader discourse on responsible AI development and public engagement in determining model behaviour.
Ericsson and Telstra have achieved a significant milestone in mobile connectivity with the launch of Ericsson’s 4th generation Radio Access Network (RAN) compute platform. The advanced technology introduces the RAN Processor 6672 in a Centralized RAN (C-RAN) configuration, offering over three times the capacity of previous models and greatly enhancing data speeds, reliability, and efficiency.
The new RAN Compute units handle critical digital signal processing tasks and achieve up to 60% lower energy consumption than traditional setups. Such efficiency supports sustainability goals while improving network performance. Additionally, the platform integrates advanced automation and AI/ML capabilities, enabling support for up to 20 times more AI models and paving the way for future technologies like 5G Advanced and 6G.
Emilio Romeo, Head of Ericsson, Australia, and New Zealand, highlighted the global importance of this deployment, noting it enhances current services and prepares the network for future innovations. Sri Amirthalingam, Telstra’s Executive for Wireless Network Engineering, emphasised that this technology will provide a superior mobile experience and support increased network capacity, setting the stage for future advancements.
Japan’s defence ministry is ramping up investments in AI, automation, and troop welfare as it tackles a growing recruitment crisis. The ministry announced these measures alongside its latest budget request, aiming to maintain a strong military presence amid a shrinking birth rate and increasing security concerns surrounding China.
A 6.9% increase in defence spending has been proposed, totalling a record 8.5 trillion yen ($59 billion). The strategy shift will fund AI surveillance systems for military bases, unmanned drones, and new automated warships with smaller crews. The move comes after Japan’s Self Defence Forces (SDF) experienced its worst recruitment year, enrolling less than half its target.
To address the shortfall, the defence ministry plans to offer financial incentives and improve living conditions, including more private sleeping quarters and better social media access. Special efforts are being made to increase female recruitment, with new accommodation and enhanced harassment prevention measures also included in the budget.
Japan is also exploring outsourcing training and support operations to civilian contractors and former SDF members, freeing up more troops for frontline duties.
As China’s military power grows, these steps form part of Prime Minister Fumio Kishida’s broader strategy to strengthen Japan’s defence capabilities.
AI startups OpenAI and Anthropic have agreed with the US Artificial Intelligence Safety Institute to collaborate on research, testing, and evaluating their advanced AI models. The deals come as regulatory scrutiny over AI’s safe and ethical development increases across the tech industry.
The agreements allow the US AI Safety Institute early access to significant new AI models from both companies before and after their release. The partnership will evaluate their capabilities and potential risks and provide feedback on safety improvements. OpenAI’s chief strategy officer, Jason Kwon, supported the initiative, citing its importance in setting a global framework for AI safety.
Anthropic, backed by Amazon and Alphabet, did not immediately comment on the deal. The US AI Safety Institute is also working closely with its counterpart in the UK to ensure international collaboration on AI safety. The institute was established as part of an executive order by President Biden to address risks associated with emerging AI technologies.
Apple and Nvidia are reportedly in discussions to invest in OpenAI, potentially pushing the valuation of the ChatGPT creator above $100 billion. The speculation of the two tech giants over the investment follows reports that venture capital firm Thrive Capital plans to invest around $1 billion in OpenAI, leading the latest funding round. While Apple and OpenAI have not commented on the news, sources indicate that Apple is increasingly integrating OpenAI’s technology into its products, including bringing ChatGPT to Apple devices earlier this year.
Microsoft, OpenAI’s largest investor with over $10 billion already committed, is also expected to join this fundraising effort. The exact amounts that Apple, Nvidia, and Microsoft will invest remain undisclosed. OpenAI’s high valuation underscores the intense competition in the AI industry, which has seen companies across various sectors invest heavily to leverage the technology and stay ahead.
The rapid rise in OpenAI’s worth reflects its pivotal role in the ongoing AI race, particularly following the launch of ChatGPT in late 2022. The company’s valuation reached $80 billion earlier this year through a tender offer led by Thrive Capital, highlighting its growing influence and strategic importance in the tech industry.
Nations worldwide are boosting demand for Nvidia’s AI chips by developing AI models tailored to their languages and cultures. Countries increasingly adopt generative AI for national security and regional needs, contributing significantly to Nvidia’s revenues. The company’s forecast predicts low double-digit billions in revenue from these AI-driven initiatives by January 2025.
Nvidia’s hardware, such as the H200 graphics processors, plays a crucial role in building AI infrastructure, with Japan‘s National Institute of Advanced Industrial Science and Technology being a notable example. These efforts highlight the importance of AI expertise and infrastructure as national priorities.
While Nvidia faces challenges due to US export controls on chip sales to China, other regions continue to drive the company’s growth. Countries aim to build AI models customised to their political, cultural, and scientific contexts, which are essential for maintaining sovereignty in an AI-driven world.
Businesses are also tapping into this trend, with firms like IBM assisting nations like Saudi Arabia in developing AI models in regional languages. Nvidia’s GPUs are expected to benefit significantly from these global efforts to build national AI platforms.
Google is weighing plans to build its first large-scale data centre in Vietnam near Ho Chi Minh City. The project, still under internal discussion, would make Google the first major US tech firm to invest in such infrastructure in the country. A decision has yet to be finalised, but the data centre could be operational by 2027.
Vietnam‘s developing digital economy and the increasing demand for cloud services and YouTube content drive Google’s interest in the region. Despite this, the country has historically struggled to attract large tech investments due to its unreliable power grid, outdated infrastructure, and less favourable regulations. However, recent reforms have opened up opportunities for foreign ownership.
Vietnam lags behind neighbouring Southeast Asian countries like Singapore, Malaysia, and Thailand, which have successfully attracted significant data centre investments. Google’s planned facility would be one of the largest in the industry, potentially consuming as much power as a small city. Estimates suggest such a centre could cost up to $650 million.
As part of its broader growth strategy in Vietnam, Google has already opened a representative office and is hiring staff locally. It also launched initiatives such as offering AI scholarships and supporting startups, reflecting its deepening commitment to the country’s digital transformation.
