Artificial intelligence

US tightens controls on China’s tech sector amid security fears

The United States has added six subsidiaries of China’s leading cloud computing firm, Inspur Group, along with dozens of other Chinese entities, to its export restriction list.

Washington accuses the companies of aiding China’s military by developing supercomputers and advanced AI technologies. The move is part of a broader strategy to curb China’s progress in high-performance computing, quantum technology, and hypersonic weapons development.

Other companies from Taiwan, Iran, Pakistan, South Africa, and the UAE were also included in the latest restrictions. China has strongly condemned the US decision, calling it an attempt to ‘weaponise trade and technology.’

The Chinese foreign ministry has vowed to take necessary measures to protect its firms, while the Beijing Academy of Artificial Intelligence, which was also targeted, called for the restrictions to be withdrawn.

Companies added to the US Entity List require special licences to access American technology, which are unlikely to be granted. The restrictions could impact major Chinese tech firms linked to AI and computing, such as Huawei and Sugon.

The United States Commerce Department argues that these measures are necessary to prevent China and other countries from using American technology for military applications. Officials insist they will not allow adversaries to strengthen their military capabilities with US-made components.

The latest crackdown follows a 2023 decision to blacklist Inspur Group, which led to scrutiny of its business ties with major US chipmakers such as Nvidia and AMD. Washington also aims to block Iran’s procurement of drone and missile technology as part of its broader national security efforts.

For more information on these topics, visit diplomacy.edu.

Google launches advanced Gemini 2.5 AI

Google has unveiled its new Gemini 2.5 AI models, starting with the experimental Gemini 2.5 Pro version.

Described as ‘thinking models’, these AI systems are designed to demonstrate advanced reasoning abilities, including the capacity to analyse information, make logical conclusions, and handle complex problems with context and nuance.

The models aim to support more intelligent, context-aware AI agents in the future.

The Gemini 2.5 models improve on the Gemini 2.0 Flash Thinking model released in December, offering an enhanced base model and better post-training capabilities.

The Gemini 2.5 Pro model, which has already been rolled out for Gemini Advanced subscribers and is available in Google AI Studio, stands out for its strong reasoning and coding skills. It excels in maths and science benchmarks and can generate fully functional video games from simple prompts.

It is also expected to handle sophisticated tasks, from coding web apps to transforming and editing code. Google’s future plans involve incorporating these ‘thinking’ capabilities into all of its AI models, aiming to enhance their ability to tackle more complex challenges in various fields.

For more information on these topics, visit diplomacy.edu.

PsiQuantum targets functional quantum machine by 2029

Quantum computing firm PsiQuantum is reportedly raising at least $750 million in a new funding round led by BlackRock, pushing the startup’s pre-money valuation to $6 billion.

The round remains ongoing, but it signals strong investor confidence in PsiQuantum’s ambitious timeline to deliver a fully functional quantum computer by 2029, or sooner.

The US, California-based company uses photonics and semiconductor techniques to produce quantum chips in partnership with GlobalFoundries at a facility in New York.

It has also secured collaborations with the governments of Australia and the US to build quantum computers in Brisbane and Chicago.

The Chicago project will anchor the new Illinois Quantum and Microelectronics Park, marking a major milestone in the commercialisation of quantum technologies.

PsiQuantum faces stiff competition from tech giants like Google, Microsoft, Amazon, and Nvidia, all of whom are making significant strides in quantum research.

For more information on these topics, visit diplomacy.edu.

AI powers Microsoft’s latest security upgrade

Microsoft has launched a new set of AI agents as part of its Security Copilot platform, aiming to automate key cybersecurity tasks like phishing detection, data protection, and identity management. The release includes six in-house agents and five developed with partners.

Among the tools is a phishing triage agent that can autonomously process routine alerts, freeing analysts to focus on advanced incidents.

Microsoft said its new AI-driven approach goes beyond traditional security platforms, using generative AI to prioritise threats, correlate data, and even recommend or execute responses.

The rollout also brings new capabilities to Microsoft Defender, Entra, and Purview, enhancing organisations’ ability to manage and secure AI systems.

While analysts welcome the move as a step forward in proactive cybersecurity, some warn that full reliance on one platform carries strategic risks like vendor lock-in and reduced flexibility.

Experts suggest a balanced approach that combines Microsoft’s core capabilities with specialised solutions for areas such as threat intelligence and cloud protection, helping organisations stay agile in a fast-evolving threat landscape.

For more information on these topics, visit diplomacy.edu.

Cerebras IPO faces further delays

Cerebras Systems’ plans for a public listing remain in limbo as a national security review by the US government continues to delay the AI chipmaker’s IPO.

The review, conducted by the Committee on Foreign Investment in the United States (CFIUS), is assessing a $335 million investment from Abu Dhabi-based AI firm G42, which has faced scrutiny over its past ties to China.

While executives had hoped for a smoother process under President Trump, delays in filling key political positions have further complicated approval.

Without clarity on G42’s stake, investors remain cautious, making it difficult for Cerebras to move forward. The situation reflects a broader reality for Wall Street, as expectations of a more deal-friendly environment under Trump have yet to materialise.

Analysts suggest that instead of rolling back Biden-era policies, the administration is likely to maintain or even expand scrutiny on foreign investments, particularly those linked to China.

Instead of a setback, Cerebras remains optimistic that the deal will be approved, with plans to proceed with its IPO once clearance is granted.

The company, valued at $8 billion last year, has seen its worth nearly double since then. Meanwhile, G42 has distanced itself from Huawei and secured a national security agreement with the US in an effort to gain regulatory approval.

For more information on these topics, visit diplomacy.edu.

AI physiotherapy service helps UK patients manage back pain

Lower back pain, one of the world’s leading causes of disability, has left hundreds of thousands of people in the UK stuck on long waiting lists for treatment. To address the crisis, the NHS is trialling a new solution: Flok Health, the first AI-powered physiotherapy clinic approved by the Care Quality Commission.

The app offers patients immediate access to personalised treatment plans through pre-recorded videos driven by artificial intelligence.

Created by former Olympic rower Finn Stevenson and tech expert Ric da Silva, Flok aims to treat straightforward cases that don’t require scans or hands-on intervention.

Patients interact with an AI-powered virtual physio, responding to questions that tailor the treatment pathway, with over a billion potential combinations. Unlike generative AI, Flok uses a more controlled system, eliminating the risk of fabricated medical advice.

The service has already launched in Scotland and is expanding across England, with ambitions to cover half the UK within a year. Flok is also adding treatment for conditions like hip and knee osteoarthritis, and women’s pelvic health.

While promising, the system depends on patients correctly following instructions, as the AI cannot monitor physical movements. Real physiotherapists are available to answer questions, but they do not provide live feedback during exercises.

Though effective for some, not all users find AI a perfect fit. Some, like the article’s author, prefer the hands-on guidance and posture corrections of human therapists.

Experts agree AI has potential to make healthcare more accessible and efficient, but caution that these tools must be rigorously evaluated, continuously monitored, and designed to support – not replace – clinical care.

For more information on these topics, visit diplomacy.edu.

OpenAI improves ChatGPT voice to stop unwanted interruptions

OpenAI has rolled out improvements to its ChatGPT voice assistant in an effort to make interactions more natural and less frustrating.

The latest update to Advanced Voice Mode allows users to pause mid-sentence without being cut off—a common issue with AI voice tools that often interrupt users during brief silences or deep breaths.

Free users of ChatGPT now benefit from this smoother conversational experience, while paying subscribers—including those on the Plus, Teams, Edu, Business, and Pro plans—also enjoy a more refined voice assistant.

The AI now speaks with a more engaging and creative tone, delivering answers that are clearer and more direct.

OpenAI’s goal is to create a more personable assistant that feels comfortable to speak with. According to the company, the changes make the AI more enjoyable to use and better equipped for real-time conversation.

These upgrades come as competition in the AI voice market heats up. Startups like Sesame, backed by major investors, are gaining attention with lifelike AI voices, while tech giants like Amazon are preparing to release enhanced voice assistants powered by large language models.

For more information on these topics, visit diplomacy.edu.

DeepSeek launches V3 to challenge OpenAI

Chinese AI startup DeepSeek has unveiled a major upgrade to its V3 large language model, intensifying the competition with US tech giants like OpenAI and Anthropic.

The new model, DeepSeek-V3-0324, is available via the AI development platform Hugging Face, showcasing significant advancements in reasoning and coding capabilities.

Benchmark tests have highlighted notable improvements in technical performance instead of its predecessor. DeepSeek, which has quickly gained recognition in the AI industry, continues to release competitive models, offering lower operational costs than many Western counterparts.

Following the V3 launch in December, DeepSeek also introduced its R1 model in January, further establishing its presence in the global AI market.

For more information on these topics, visit diplomacy.edu.

Canada warns of foreign election interference

Canada’s intelligence agency has warned that China and India are highly likely to interfere in the country’s general election on 28 April, with Russia and Pakistan also having the potential to do so.

The Canadian Security Intelligence Service (CSIS) stated that while previous interference attempts in the 2019 and 2021 elections did not alter the results, the country had been slow to respond at the time. Both China and India have denied previous allegations of meddling in Canada’s internal affairs.

Vanessa Lloyd, CSIS’s deputy director of operations, said hostile states are increasingly using AI to influence elections, with China being particularly likely to exploit such tools.

The warning comes amid tense diplomatic relations between Canada and Beijing, following China’s recent tariffs on $2.6 billion worth of Canadian agricultural products and Ottawa’s strong condemnation of China’s execution of four Canadian citizens on drug charges.

India has also been under scrutiny, with Canada expelling six Indian diplomats last year over allegations of involvement in a plot against Sikh separatists.

Lloyd stated that India has both the intent and capability to interfere in Canadian politics and communities, though the Indian diplomatic mission in Ottawa has yet to comment.

She added that while it is difficult to directly link foreign interference with election outcomes, such activities undermine public trust in Canada’s democratic institutions.

For more information on these topics, visit diplomacy.edu.

Instagram users react to Meta’s new AI experiment

Meta has come under fire once again, this time over a new AI experiment on Instagram that suggests comments for users. Some users accused the company of using AI to inflate engagement metrics, potentially misleading advertisers and diminishing authentic user interaction.

The feature, spotted by test users, involves a pencil icon next to the comment bar on Instagram posts. Tapping it generates suggested replies based on the image’s content.

Meta has confirmed the feature is in testing but did not reveal plans for a broader launch. The company stated that it is exploring ways to incorporate Meta AI across different parts of its apps, including feeds, comments, groups, and search.

Public reaction has been largely negative, with concerns that AI-generated comments could flood the platform with inauthentic conversations. Social media users voiced fears of fake interactions replacing genuine ones, and some accused Meta of deceiving advertisers through inflated statistics.

Comparisons to dystopian scenarios were common, as users questioned the future of online social spaces.

This isn’t the first time Meta has faced backlash for its AI ventures. Previous attempts included AI personas modelled on celebrities and diverse identities, which were criticised for being disingenuous and engineered by largely homogenous development teams.

The future of AI-generated comments on Instagram remains uncertain as scrutiny continues to mount.

For more information on these topics, visit diplomacy.edu.