Artificial intelligence

Musk denies xAI-Tesla collaboration claims on AI technology

The CEO of Tesla has denied claims that his AI startup xAI had entered discussions to share future Tesla revenue in exchange for giving the automaker access to its technology. The Wall Street Journal reported that Tesla was considering licensing xAI’s artificial intelligence models to enhance its full self-driving software and splitting revenue with the startup.

Musk refuted the report, stating that although Tesla had benefited from conversations with xAI engineers, there was no need to license any technology. He called the article ‘not accurate’ in a post on social media platform X.

The Journal’s report suggested that xAI could also help Tesla develop other features, such as a voice assistant for its electric vehicles and software for its humanoid robot, Optimus. Musk has previously mentioned that xAI could play a role in advancing Tesla’s self-driving capabilities and building a new data centre.

The billionaire launched xAI last year to rival OpenAI, with plans to integrate its AI chatbot, Grok, into Tesla’s systems. Discussions have reportedly taken place regarding a potential $5 billion investment in xAI by Tesla.

Egypt Prime Minister secures key tech and telecom MoUs with China

Egypt Prime Minister Mostafa Madbouly signed five key Memoranda of Understanding (MoUs) with Chinese firms and institutions to enhance Egypt-China telecommunications and information technology cooperation. These agreements, made during the Forum on China-Africa Cooperation (FOCAC) in Beijing, mark a significant development in Egypt’s tech and infrastructure sectors.

The first MoU with FiberHome Telecommunication Technologies involves setting up a fibre optic cable factory in Egypt, producing one million fibre kilometres annually and creating 200 jobs. It will also include a research and development centre and a training facility for network engineers.

The second MoU, with ITIDA, Tsinghua Unigroup, and Telecom Egypt, focuses on building a data centre and cloud services operation supported by a $300 million investment fund. This partnership will also establish a research centre for semiconductor design and develop AI applications, including an Arabic language model.

Huawei Egypt’s MoU will establish a development centre for local industry solutions, software, and cloud computing, aiming to train 1,500 developers by 2025 and support startups with cloud resources. The fourth MoU with ZTE will localise network equipment production and establish training labs for 5G and GPON technologies, providing training for 1,200 participants.

The final MoU with Hengtong Group will create a second fibre optic cable factory in the Suez Canal Economic Zone with a $15 million investment, producing 3 million kilometres of cables annually and including a training academy in collaboration with the National Telecommunications Institute. These agreements highlight Egypt’s commitment to advancing its technological infrastructure and deepening its partnership with China.

US proposes mandatory reporting for advanced AI and cloud providers

The US Commerce Department has proposed new rules that would require developers of advanced AI and cloud computing providers to report their activities to the government. The proposal aims to ensure that cutting-edge AI technologies are safe and secure, particularly against cyberattacks.

It also mandates detailed reporting on cybersecurity measures and the results of ‘red-teaming’ efforts, where systems are tested for vulnerabilities, including potential misuse for cyberattacks or the development of dangerous weapons.

The move comes as AI, especially generative models, has sparked excitement and concern, with fears over job displacement, election interference, and catastrophic risks. Under the proposal, the collected data would help the US government enforce safety standards and protect against threats from foreign adversaries.

Why does this matter?

The following regulatory push follows President Biden’s 2023 executive order requiring AI developers to share safety test results with the government before releasing certain systems to the public. The new rules come amid stalled legislative action on AI and are part of broader efforts to limit the use of US technology by foreign powers, particularly China.

South Korea hosts global summit on AI in warfare

South Korea hosted a pivotal international summit on Monday to craft guidelines for the responsible use of AI in the military. Representatives from over 90 countries, including the US and China, attended the two-day event in Seoul. The summit aimed to produce a blueprint for AI use in warfare, though any agreement is expected to lack binding legal power. The initiative marked the second such gathering, following a similar summit in Amsterdam last year, where nations endorsed a call to action without legal obligations.

South Korean Defense Minister Kim Yong-hyun highlighted AI’s growing role in modern warfare, referencing Ukraine’s use of AI-powered drones in its ongoing conflict with Russia. He likened AI’s potential in the military to a ‘double-edged sword,’ emphasising its ability to enhance operational capabilities and its risks if misused. South Korea‘s foreign minister, Cho Tae-yul, further underscored the need for international safeguards, suggesting that mechanisms be put in place to prevent autonomous weapons from making lethal decisions without human oversight.

The summit aims to outline principles for the responsible use of AI in the military, drawing from guidelines established by NATO and various national governments. However, many attending nations will endorse the proposed frame, which remains to be seen. While the document seeks to establish minimum guardrails for AI, it is not expected to impose legally binding commitments.

Beyond this summit, international discussions on AI’s role in warfare are ongoing. ThUN also explores potential restrictions on lethal autonomous weapons under the 1983 Convention on Certain Conventional Weapons (CCW). Additionally, the US government has been leading efforts to promote responsible AI use in the military, with 55 countries already endorsing its declaration.

Co-hosted by the Netherlands, Singapore, Kenya, and the United Kingdom, Seoul brings together around 2,000 participants, including representatives from international organisations, academia, and the private sector, discussing various topics, from civilian protection to AI’s potential role in nuclear weapon control. The summit seeks to ensure ongoing collaboration on the rapidly evolving technology, especially as governments remain the key decision-makers in this crucial area.

FedEx expands fulfilment with investment in AI robotics firm Nimble

FedEx has made a strategic investment in AI robotics and automation company Nimble to enhance its fulfilment services for small and medium-sized businesses. The investment aims to support FedEx’s Fulfilment unit, which assists businesses with order fulfilment and inventory management.

The investment comes as parcel delivery companies increasingly turn to automation to reduce costs and improve efficiency, particularly during periods of lower freight demand. FedEx believes Nimble’s automated third-party logistics solutions will help optimise supply chain operations across North America.

Scott Temple, president of FedEx Supply Chain, stated that the alliance with Nimble will expand the company’s presence in e-commerce, allowing FedEx to scale its fulfilment offerings throughout North America. The exact size of the investment has not been disclosed.

Nimble’s AI robotics technology is expected to help FedEx improve the efficiency of its fulfilment operations and further strengthen its position in the e-commerce sector.

ChatGPT gains over million subscribers, new pricing plans discussed

OpenAI announced on Thursday that it now has over 1 million paying users across its ChatGPT business products, including Enterprise, Team, and Edu. The increase from 600,000 users in April highlights CEO Sam Altman’s success in driving enterprise adoption of the AI tool.

Recent reports suggest OpenAI executives are discussing premium subscriptions for upcoming large-language models, such as the reasoning-focused Strawberry and a new flagship model called Orion. Subscription prices could reach as high as $2,000 per month for these advanced AI tools.

ChatGPT Plus currently costs $20 per month, while the free tier continues to be used by hundreds of millions every month. OpenAI is also working on Strawberry to enable its AI models to perform deep research, refining them after their initial training.

The discussion around premium pricing follows news that Apple and Nvidia are in talks to invest in OpenAI, with the AI company expected to be valued at over $100 billion. ChatGPT currently has more than 200 million weekly active users, doubling its user base since last autumn.

NIST releases new digital identity and AI guidelines for contractors

US National Institute of Standards and Technology (NIST) has released a new draft of its Digital Identity Guidelines, introducing updates for government contractors in cybersecurity, identity verification, and AI use. The guidelines propose expanded identity proofing methods, including remote and onsite verification options. These enhancements aim to improve the reliability of identity systems used by government contractors to access federally controlled facilities and information. By providing different assurance levels for identity verification, NIST ensures that contractors can implement secure and appropriate measures based on the context and location of the verification process.

A significant focus of the guidelines is on continuous evaluation and monitoring. Organisations are now required to implement ongoing programs that track the performance of identity management systems and evaluate their effectiveness against emerging threats. The guidelines also emphasise the importance of proactive fraud detection. Contractors and credential service providers (CSPs) must continuously assess and update their fraud detection methods to align with the evolving threat landscape.

One of the notable updates in the guidelines is the introduction of syncable authenticators and digital wallets. This allows contractors to manage their digital credentials more efficiently by storing them securely in digital wallets. These wallets provide flexibility in how contractors present their identity attributes when accessing different federal systems.

The guidelines also introduce a risk-based approach to authentication, where authentication levels are tailored to the sensitivity of the system or information being accessed. That gives government agencies the flexibility to assign different authentication methods depending on the security needs of the transaction. For example, accessing highly sensitive systems would require stronger multi-factor authentication (MFA) measures, including biometrics, while less critical systems may have less stringent requirements.

Why does this matter?

The use of AI and ML in identity systems is another key aspect of the Draft Guidelines. NIST emphasises transparency and accountability in integrating AI and ML into these systems. Organisations must document how AI is used, disclose the datasets for training models, and ensure that AI systems are evaluated for risks like bias and inequitable outcomes. The guidelines address the concern that AI technologies could exacerbate existing inequities or produce biassed results in identity verification processes. Organisations are encouraged to adopt NIST’s AI Risk Management Framework to mitigate these risks and consult its guidance on managing bias in AI.

Lastly, the guidelines highlight the importance of privacy, equity, and usability in digital identity systems. Ensuring broad participation and access to digital services, especially for individuals with disabilities, is a core requirement. NIST stresses that digital identity systems must be designed to be inclusive and accessible to all contractors, addressing any potential usability challenges while maintaining security.

Qualcomm explores buying parts of Intel’s business

Qualcomm is exploring the possibility of acquiring parts of Intel’s design business to enhance its product lineup, according to sources familiar with the discussions. The mobile chipmaker is particularly interested in Intel’s client PC design business but also considers other segments. Intel, currently facing financial difficulties and seeking to offload assets, has not been approached by Qualcomm and has stated its commitment to its PC business.

Qualcomm, known for its smartphone chips and a major Apple customer, has been contemplating the acquisition for several months. However, the company’s plans remain uncertain and could evolve. Intel recently reported a significant drop in PC business revenue and has implemented cost-cutting measures, including staff reductions and a dividend pause, as it struggles to fund its manufacturing operations.

Intel’s board will meet next week to review options for trimming operations, which may include selling its programmable chip unit, Altera. Meanwhile, Intel recently launched a new AI-focused PC chip called Lunar Lake, signalling its continued push to innovate despite financial challenges.

Former OpenAI scientist aims to develop superintelligent AI safely

Ilya Sutskever, OpenAI’s former chief scientist, has launched a new company called Safe Superintelligence (SSI) to develop safe AI systems that significantly surpass human intelligence. In an interview, Sutskever explained that SSI aims to take a different approach to AI scaling compared to OpenAI, emphasising the need for safety in superintelligent systems. He believes that once superintelligence is achieved, it will transform our understanding of AI and introduce new challenges for ensuring its safe use.

Sutskever acknowledged that defining what constitutes ‘safe’ AI is still a work in progress, requiring significant research to address the complexities involved. He also highlighted that as AI becomes more powerful, safety concerns will intensify, making it essential to test and evaluate AI systems rigorously. While the company does not plan to open-source all of its work, there may be opportunities to share parts of its research related to superintelligence safety.

SSI aims to contribute to the broader AI community’s safety efforts, which Sutskever views positively. He believes that as AI companies progress, they will realise the gravity of the safety challenges they face and that SSI can make a valuable contribution to this ongoing conversation.

Global AI framework signed to safeguard human rights

The UK has become one of the first signatories of an international treaty designed to regulate AI and prevent its misuse. This legally binding agreement, drafted by the Council of Europe and signed by countries including the EU, US, and Israel, mandates safeguards to protect human rights, democracy, and the rule of law from potential AI threats. Governments are expected to tackle risks such as AI-generated misinformation and the use of biassed data in decision-making processes.

The treaty outlines several key principles, including ensuring data protection, non-discrimination, and the responsible development of AI. Both public and private sector AI users will be required to assess the impact of AI systems on human rights and provide transparency to the public. Individuals will also have the right to challenge AI-made decisions and file complaints with relevant authorities, ensuring accountability and fairness in AI applications.

In the UK, the government is reviewing how to implement the treaty’s provisions within existing legal frameworks, such as human rights laws. A consultation on a new AI bill is underway, which could further strengthen these safeguards. Once ratified, the treaty will allow authorities to impose sanctions, including bans on certain AI uses, like systems utilising facial recognition from unauthorised data sources.

Diplo chatbot can make mistakes. Consider checking important information.