Artificial intelligence

OpenAI to introduce content creator control in AI development

OpenAI has announced that it’s developing a tool to enhance ethical content usage for AI development. The tool, called Media Manager, allows content creators to specify the use of their work in AI training, aligning with the digital rights movement and addressing long-standing issues around content usage in a context where it faces a growing number of copyright infringement lawsuits.

The concept isn’t entirely new. It parallels the decades-old robots.txt standard used by web publishers to control crawler access to website content. Last summer, OpenAI adapted this idea, pioneering the use of similar permissions for AI, thus allowing publishers to set preferences for the use of their online content in AI model training.

However, many content creators do not control the websites where their content appears, and their work is often used in various forms across the internet, rendering previously proposed solutions as nonsufficient. The Media Manager tool comes as an attempt to create a more scalable and efficient solution for creators to assert control over their content’s use in AI systems. It is being developed as a comprehensive tool that will allow creators to register their content and specify inclusion or exclusion from AI research and training. OpenAI plans to enhance this tool over time with more features, supporting a broader range of creator needs. This initiative involves complex machine learning research to develop a system capable of identifying copyrighted text, images, audio, and video across diverse sources. 

OpenAI is collaborating with creators, content owners, and regulators to shape the Media Manager tool, with an expected launch by 2025. This collaborative approach aims to develop the tool in a way that meets the nuanced requirements of various stakeholders and sets a standard for the AI industry.

Why does it matter?

The significance of OpenAI’s Media Manager stems from its attempt to respond to the foundational discord of how AI interacts with human-generated content. By providing tools that respect and enforce the rights of creators, OpenAI is fostering a sustainable model where AI development is aligned with ethical and legal standards. This initiative is crucial for ensuring that AI technologies are developed in ways that do not exploit but instead respect and contribute positively to the creative economy. It sets a precedent for transparency and responsibility that could influence the entire AI industry towards more ethical practices.

Nvidia and Databricks sued for alleged copyright infringement in AI model development

Nvidia Corporation and Databricks Inc. face class-action lawsuits alleging copyright infringement in the creation of their AI models. The litigation highlights a growing concern over the use of copyrighted content without permission.

The lawsuits against Nvidia corp. and Databricks Inc., filed on March 8 by authors Abdi Nazemian, Brian Keene, and Stewart O’Nan in the U.S. District Court for the Northern District of California, argue that Nvidia’s NeMo Megatron and Databricks’ MosaicML models were trained on vast datasets containing millions of copyrighted works. Notably, the complaints suggest these models include content from well-known authors like Andre Dubus III and Susan Orlean, among others, without their consent. This has sparked a broader debate on whether such practices constitute fair use, as AI developers claim, or if they infringe upon the copyrights of individual creators.

The core of the dispute lies in how AI companies compile their training data. Reports indicate that some of the data used included copyrighted material from ‘shadow libraries’ like Bibliotik, which hosts and distributes unlicensed copies of nearly 200,000 books. The involvement of such sources in training datasets could potentially undermine the legality of the AI training process, which relies on the ingestion of large volumes of text to produce sophisticated AI outputs.

Legal experts and industry analysts are closely watching these cases, as the outcomes could set important precedents for the future of AI development. Companies like Nvidia have defended their practices, stating that their development processes comply with copyright laws and emphasizing the transformative nature of AI technology. However, the plaintiffs argue that this does not justify the unauthorized use of their work, which they claim undermines their financial and creative rights.

The lawsuits against Nvidia and Databricks are part of a larger trend of legal challenges that tech giants face regarding their development of AI technologies and using copyrighted materials to train large language models (LLMs), designed to process and generate human-like text.

OpenAI, the creator of the AI model known as ChatGPT, faced similar legal scrutiny when the New York Times filed a lawsuit against it, alleging that the company used copyrighted articles to train its language models without permission. 

These developments raise crucial questions about the balance between innovation and copyright protection in the digital context.

House of Lords committee criticizes UK government for inaction on copyright issues with LLMs

The UK government faces increasing scrutiny over its handling of copyright issues related to the development of large language models (LLMs). After an extensive inquiry, the House of Lords Communications and Digital Committee issued a report which highlighted a critical need for clearer regulations to protect content creators and prevent misuse of copyrighted materials in AI training processes.

The House of Lords has expressed dissatisfaction with the government’s current efforts, describing them as “inadequate and deteriorating.” The Committee’s latest report emphasizes the need for a definitive governmental stance on copyright application to LLMs, urging for legislative updates to ensure the law remains effective in the context of digital change.

The Committee’s concerns are not isolated. Various stakeholders, including industry experts and intellectual property organizations, have voiced similar frustrations. For instance, Getty Images criticized the prevalent ‘ask for forgiveness later’ approach, which is contrary to fundamental copyright principles that require prior permission from rights holders. The lack of a clear government position has led to a contentious environment where larger publishers might secure licensing deals, potentially sidelining smaller content creators.

The government’s response, detailed in their official statement, points to ongoing efforts to introduce a transparency mechanism for copyright holders and potential legislative updates. However, this response has been criticized as insufficient, with accusations that it favors large tech firms, allowing them to entrench business models potentially harmful to the UK’s creative industries.

Furthermore, the Communications and Digital Committee has proposed measures to promote fair competition and transparency in the AI market, stressing the importance of preventing market dominance by a few large companies. They advocate for a balanced approach that does not solely focus on AI safety but also considers AI technologies’ economic and societal impacts.

This ongoing debate highlights the complex interplay between innovation, regulation, and protection of intellectual property amid the ongoing digital transformation.

Artists take legal action against Google: copyright battle over Imagen AI training

A legal battle is unfolding as visual artists take on Google and its parent company Alphabet, over the use of their copyrighted works without consent to train their AI image generator, Imagen. The case, brought forth by prominent artists including Jingna Zhang, Sarah Andersen, Hope Larson, and Jessica Fink, alleges that Google’s Imagen AI used their copyrighted artwork as part of its training data without permission. This lawsuit highlights a significant copyright issue for AI development, where training datasets often contain millions of images scraped from various sources, potentially including copyrighted material.

The artists are seeking legal remedies, including damages and the destruction of all copies of their works allegedly made during Imagen’s training. This legal action echoes a broader concern among creative professionals about the implications of AI on intellectual property rights. 

The case’s progression through the judicial system is being closely watched, as it may set significant precedents for future AI copyright disputes and comes as one of many similar AI copyright infringement lawsuits:

  • Getty Images vs. Stability AI: Getty Images sued Stability AI in both the US and UK for using millions of its copyrighted images to train the AI model Stable Diffusion without permission. 
  • Visual Artists vs. Stability AI, Midjourney, and DeviantArt: A class action by visual artists against AI developers for using copyrighted artworks to train AI models without consent. The artists argue that the AI-generated images are derivative works, raising questions about the boundaries of originality and reproduction in AI outputs.
  • Programmers vs. GitHub and OpenAI: Programmers filed a lawsuit against GitHub, Microsoft, and OpenAI alleging that their code was used without proper licensing to train AI models.

The U.S. Copyright Office has actively sought public input on how to handle copyright infringements involving AI. This includes discussions on whether there’s enough human involvement in AI outputs to consider them original creations by humans or if they are mere reproductions of scraped data .

Major tech firms argue that their use of copyrighted content in AI training is a form of “fair use” aimed at identifying patterns rather than replicating the original works. This stance is part of a broader industry argument that AI’s training processes do not specifically aim to extract or reproduce individual copyrighted works but rather to learn from a broad corpus of data to generate new content. 

As the legal battles unfold, the outcomes will likely influence how copyright laws are interpreted in the context of AI. The discussions around these cases reflect a growing recognition of the need for updated legal frameworks that adequately address the nuances of AI technology and its impact on creative industries.

This ongoing situation presents a critical juncture for the intersection of technology and copyright law, offering an opportunity to redefine the boundaries of authorship and intellectual property in the digital context.

Vidu: China’s competitor to OpenAI’s Sora in AI video generation

Tsinghua University in collaboration with Shengshu Technology has unveiled ‘Vidu,’ a new AI tool capable of generating short videos from text prompts, positioning itself as a competitor to OpenAI’s Sora.  

Vidu is designed to generate 1080p quality videos lasting up to 16 seconds, in contrast to Sora, which produces videos up to 60 seconds long. However, it is tailored specifically to incorporate elements of Chinese culture into its outputs. This feature distinctly positions Vidu as a cultural venture in addition to it being a technological innovation. During the announcement at the Zhongguancun Forum in Beijing, Zhu Jun, Chief Scientist at Shengshu Technology and Deputy Dean at Tsinghua’s Institute for AI, highlighted Vidu’s unique features. According to Beijing News, he described Vidu as ‘imaginative,’ capable of simulating the physical world and producing 16-second videos with consistent characters, scenes, and timelines’ that comprehend and integrate Chinese elements. The launch event showcased Vidu’s capability to interpret and visually represent text inputs into engaging video clips with vivid details.

This initiative is part of a broader strategy by China to develop cutting-edge technologies, particularly generative AI, to compete on the global stage. The announcement of Vidu comes at a time when AI-generated media is gaining significant attention worldwide, with applications ranging from entertainment to educational content creation. The ability of tools like Vidu to produce high-quality video content rapidly opens up new possibilities for content creators and industries looking to leverage AI for creative expression and communication.

Industry experts, however, have pointed out insufficient computing power as a significant barrier to the progress of Chinese firms like the developers of Vidu. For comparison, OpenAI’s Sora requires eight Nvidia A100 GPUs, running for over three hours, to produce a one-minute video clip. “Sora demands extensive computing resources for inferencing,” says Li Yangwei, a Beijing-based technical consultant specializing in intelligent computing. Further complicating the situation, the US has tightened export controls on advanced chips, including Nvidia’s A100 and H100 GPUs. These components are crucial for training AI systems but are now restricted from being shipped to China. This poses additional challenges to the development and deployment of advanced AI technologies like Vidu, reflecting the ongoing geopolitical tensions that impact the global technology landscape.

Financial Times partners with OpenAI to license content and develop AI tools

The Financial Times has announced a collaboration with OpenAI, allowing the AI company to license its content and utilise it to develop AI tools. Under this partnership, ChatGPT users will encounter summaries, quotes, and article links from the Financial Times, with all information sourced attributed to the publication. In return, OpenAI will collaborate with the Financial Times to innovate and create new AI products, building upon their existing relationship, as the publication already utilises OpenAI’s ChatGPT Enterprise.

John Ridding, CEO of the Financial Times Group, emphasises the importance of maintaining ‘human journalism’ even amidst collaborations with AI platforms. Ridding asserts that AI products must incorporate reliable sources, highlighting the significance of partnerships like the one with OpenAI. Notably, OpenAI has secured similar agreements with other news organisations, including Axel Springer and the Associated Press, to license content for training AI models.

However, OpenAI’s licensing agreements have drawn attention for their comparatively lower payouts to publishers, ranging from $1 million to $5 million, in contrast to offers from companies like Apple. This discrepancy has led to legal disputes, with the New York Times and other news outlets suing OpenAI and Microsoft for alleged copyright infringement related to ChatGPT’s use of their content. These legal battles underscore the complexities and challenges surrounding the integration of AI technology within the news industry.

Emerging challenges and diminished attendance at the second global AI safety summit

The second Global AI Safety Summit, scheduled for May 21-22 in Seoul, South Korea, faces a significant decrease in attendance compared to its inaugural session. Co-hosted virtually by Britain and South Korea, this year’s summit will address crucial issues surrounding AI’s impact on society and its regulatory challenges.

The inaugural summit at Bletchley Park, notable for the ‘Bletchley Declaration‘, symbolized a global commitment to AI safety and regulation, where industrial and geopolitical tensions gave way to concensus and shared stakes. This year, however, key figures have opted not to attend, turning down the invitation. Key attendees like the EU’s top tech regulators and the governments of Canada, the Netherlands, and Brazil have declined or are still considering their invitations. The lower turnout, however, may be linked to the shift in the general treatment of AI technology from the initial hype of its potential to raising questions about its limitations, according to some experts.

Expert Martha Bennett from Forrester Research emphasizes the difficulty in expanding upon the broad agreements made at Bletchley Park due to differing global perspectives on AI regulation. 

Francine Bennett from the Ada Lovelace Institute points out that ‘The policy discourse around AI has expanded to include other important concerns, such as market concentration and environmental impacts.’

While some industry giants rely on an energy breakthrough and heavy financing to boost production, Professor Jack Stilgoe, an expert in technology policy at University College London, warns that ‘The failure of the technology to live up to the hype is inevitable.’ suggesting a more measured approach to expectations and ventures in AI development.

AI governance in Italy: National strategy and law on the horizon

Italy is stepping forward with an initiative to shape the future of artificial intelligence (AI) within its borders through the Italian Strategy on AI for 2024–2026, alongside a draft law on AI. These efforts aim to ensure that AI development and use are transparent and beneficial to society, in line with global standards like the OECD AI Principles, which emphasise respect for human autonomy and decision-making.

The strategy and draft law, prepared by Digital Italy (AGID) and the Department for Digital Transformation, target four main sectors: Scientific Research, Public Administration, Enterprises, and Education: 

Scientific Research: The strategy aims to enhance the AI research ecosystem, fostering collaboration between universities, research centers, and ICT companies. It also focuses on attracting and retaining talent and financing innovative AI projects.

Public Administration: Initiatives here aim to standardise the adoption of AI in public services, improve process efficiency, and promote AI literacy among public employees.

Enterprises: The strategy supports the creation of an AI-driven ecosystem for SMEs, providing financial incentives and fostering conditions conducive to innovation and collaboration between businesses and research institutions.

Education: Efforts in this sector focus on integrating AI education into schools and universities, promoting advanced degree programs, and supporting high-level internships and training programs.

This comprehensive approach seeks to integrate AI into the Italian economy and public services effectively, promoting both innovation and ethical standards.

Prime Minister Giorgia Meloni has indicated in a statement that addressing the risks associated with artificial intelligence will be a key theme during Italy’s G7 presidency. 

The draft AI law in Italy addresses several key legal areas to adapt to the challenges posed by AI technologies. Specifically, it includes updates to data protection laws to better safeguard personal information in the context of AI usage, and introduces new provisions in criminal law to handle AI-related offenses effectively. Additionally, the law proposes changes to civil procedures to ensure they remain relevant in the digital age, and revises media laws to address the challenges of AI-generated content. These changes aim to protect citizens’ rights and promote a safe and innovative environment for AI development.

Elon Musk’s xAI aims to raise $6 billion in fundraising round

xAI, Elon Musk’s challenger to OpenAI, is amid a substantial fundraising round. According to sources familiar with the deal, it aims to raise $6 billion on an $18 billion valuation. Originally planned at $3 billion on a $15 billion valuation, the increase reflects heightened investor interest in the venture.

Investors in this round include notable names like Sequoia Capital, Future Ventures, and likely, Valor Equity Partners and Gigafund, all deeply connected to Musk’s network. The fundraising process, overseen by Jared Birchall from Musk’s family office, reportedly involved direct discussions with Musk and his engineers, showcasing the company’s intimate approach.

xAI’s vision encompasses bridging the digital and physical realms by leveraging data from Musk’s companies, including Tesla, SpaceX, Boring Company, and Neuralink. Musk’s overarching plan involves deploying AI-driven solutions, like the chatbot Grok, across his ecosystem, with future applications potentially extending to Tesla’s humanoid robot, Optimus.

Why does it matter?

For Musk, xAI’s success holds implications beyond technological advancement. With X, Musk’s troubled social media platform, holding a stake in xAI, the former benefits from the latter’s growth. Meanwhile, Musk’s ongoing feud with OpenAI, the AI giant he co-founded but later distanced himself from, adds another layer of complexity to the evolving landscape of AI.

Japan tightens export rules for advanced tech to prevent military use

The Japanese government is taking steps to tighten regulations on exporting advanced technologies, particularly those with potential military applications in countries like Russia and China. This strategic move comes amidst growing concerns in Japan about diversifying cutting-edge technologies for military purposes. The Ministry of Economy, Trade and Industry will require private firms to provide advance notification before exporting technologies such as quantum science to prevent their misuse in military endeavours.

Export controls will be expanded to cover cutting-edge technologies and non-cutting-edge fields that pose risks of being used in conventional weapons. The proposed amendment to the trade ministry ordinance will introduce penalties for violations, reinforcing existing regulations under the foreign exchange and foreign trade law. This includes strengthening the ‘catch-all regulation,’ which covers items not explicitly listed but deemed at risk of being diverted for weapons manufacturing.

Under the new regulations, firms exporting advanced technologies must notify the ministry of their intentions and undergo risk assessments. If concerns arise about the potential military diversion of the exported goods, permission from the ministry will be required. Furthermore, in response to evolving global security dynamics, even goods and technologies in fields not considered cutting-edge will be subject to approval if investigations suggest they could be repurposed for military use, as demonstrated by Russia’s recent actions in Ukraine.

To facilitate compliance, the ministry will provide firms with information on trading partners and regions posing risks related to military diversion. Additionally, clear criteria will be published to help firms assess the risk level of their transactions, reducing confusion and ensuring compliance with the law. These measures reflect Japan’s commitment to enhancing export controls and safeguarding the responsible transfer of advanced technologies in a rapidly evolving global security landscape.