Privacy and data protection

Australia’s social media laws face criticism over YouTube exemption

Australia’s government recently passed laws banning social media access for children under 16, targeting platforms like TikTok, Snapchat, Instagram, Facebook, and X. However, YouTube was granted an exemption, with the government arguing that it serves as a valuable educational tool and is not a ‘core social media application.’ That decision followed input from company executives and educational content creators, who argued that YouTube is essential for learning and information-sharing. While the government claims broad community support for the exemption, some experts believe this undermines the goal of protecting children from harmful online content.

Mental health and extremism experts have raised concerns that YouTube exposes young users to dangerous material, including violent, extremist, and addictive content. Despite being exempted from the ban, YouTube has been criticised for its algorithm, which researchers say can promote far-right ideologies, misogyny, and conspiracy theories to minors. Studies conducted by academics have shown that the platform delivers problematic content within minutes of search queries, including harmful videos on topics like sex, COVID-19, and European history.

To test these claims, Reuters created child accounts and found that searches led to content promoting extremism and hate speech. Although YouTube removed some flagged videos, others remain on the platform. YouTube stated that it is actively working to improve its content moderation systems and that it has removed content violating its policies. However, critics argue that the platform’s algorithm still allows harmful content to thrive, especially among younger users.

Google introduces new Gemini AI model

Google has quietly launched its latest AI model, Gemini 2.0 Pro Experimental, through a changelog update for its Gemini chatbot app. The new model, now available to Gemini Advanced subscribers, promises improved accuracy and performance, particularly in coding and mathematics-related tasks.

The release comes as competition in the AI space intensifies, with Chinese startup DeepSeek gaining attention for its high-performing models. Unlike previous Gemini versions, Gemini 2.0 Pro Experimental does not support real-time data access and may exhibit unexpected behaviours as it remains in an early preview phase.

Alongside this update, Google has also rolled out its Gemini 2.0 Flash model to all users of the Gemini app. The company continues to iterate rapidly on its AI technology, positioning Gemini as a key player in the evolving AI market.

Siri upgrade brings expanded language support

Apple has announced that its AI suite, Apple Intelligence, will support additional languages starting in April, including French, German, Italian, Portuguese, Spanish, Japanese, Korean, and simplified Chinese. The update will also introduce localised English versions for India and Singapore, broadening access to the technology beyond its initial US English release.

The expansion follows a December update that brought support for various English dialects, including those used in Australia, Canada, New Zealand, South Africa, and the UK. However, Apple has yet to confirm when its AI suite will be available in the EU or mainland China.

CEO Tim Cook also revealed that the next version of Siri, which will feature improved on-screen contextual understanding, is expected to launch in the coming months. The update marks Apple’s latest effort to strengthen its AI ecosystem and compete with rivals in the artificial intelligence space.

South Korea to question DeepSeek over data privacy

South Korea’s privacy watchdog plans to investigate how DeepSeek manages users’ personal data. The Personal Information Protection Commission intends to send a written request for details to the Chinese AI model’s operators.

Authorities in several other countries, including France, Italy and Ireland, have also been examining DeepSeek’s data practices. Concerns have grown over how AI models collect and process personal information.

An official from South Korea’s privacy commission confirmed that the request for information could be submitted as early as Friday. No further details were provided on the scope of the inquiry.

Taiwan restricts DeepSeek AI over security risks

Taiwan’s digital ministry has instructed government departments not to use DeepSeek’s artificial intelligence service due to security concerns. The warning highlights ongoing fears about Chinese technology, given Beijing’s claims over Taiwan and its political and military threats.

Authorities cited potential risks from cross-border data transmission and information leaks, describing DeepSeek as a threat to national security. The ministry pledged to monitor technological developments and adjust security policies accordingly.

South Korea’s privacy watchdog is also investigating DeepSeek’s handling of personal data. European regulators in France, Italy, and Ireland have launched similar inquiries into its data practices.

DeepSeek’s AI assistant recently surpassed ChatGPT in downloads from Apple’s App Store. Global investors responded by selling US tech stocks, causing Nvidia’s market value to drop by $593 billion in a historic single-day loss.

Trump and Nvidia CEO met to discuss DeepSeek and AI chip export

In a meeting at the White House on Friday, US President Donald Trump and Nvidia CEO Jensen Huang discussed the emerging challenges posed by China’s AI advancements, particularly the rapid rise of DeepSeek, a Chinese AI company that has disrupted the global tech industry. The conversation focused on strengthening US leadership in AI technology and tightening restrictions on exporting advanced AI chips to China. While President Trump did not disclose specific details, he described the meeting as ‘good’ and praised Huang as a ‘gentleman.’

DeepSeek has recently sent shockwaves through the tech world by introducing an AI model that matches the performance of leading US technologies but at a fraction of the cost. Its launch has raised concerns about China’s ability to close the gap in AI development with the US. Within days of its debut, DeepSeek became the most downloaded app in Apple’s App Store, and its emergence wiped approximately $1 trillion off the market value of US tech stocks, including a 17% plunge in Nvidia shares. The development has reignited fears that China could challenge the United States’ dominance in AI.

As part of the US response, Trump’s administration is considering further restrictions on Nvidia’s H20 chips, designed for the Chinese market. These chips, capable of powering AI software, were specifically developed to comply with previous export limits imposed by the Biden administration. However, the Trump administration aims to tighten the rules further to ensure advanced computing power stays within US borders and allied nations. Discussions among Trump officials about restricting the shipments of these chips are in the early stages, but they reflect a growing bipartisan concern about maintaining technological supremacy over China.

The president’s meeting with Nvidia’s CEO highlighted a potential opportunity for US tech companies. According to a source familiar with the discussions, Trump views DeepSeek’s ability to produce low-cost AI models as a wake-up call, suggesting that American firms could adopt more cost-efficient strategies to compete without massive spending. The administration sees this as a chance to recalibrate the US AI innovation and leadership approach.

Lawmakers on both sides of the aisle have echoed calls for tighter controls. Republican John Moolenaar and Democrat Raja Krishnamoorthi, co-leads of the House Select Committee on China, urged the Commerce and State Departments to review US export controls in light of China’s advancements. Their concerns are amplified by reports suggesting that DeepSeek may have used restricted US chips in its operations, prompting an investigation by the Commerce Department.

The US government has long relied on export controls to maintain its technological edge, but DeepSeek’s rise has exposed vulnerabilities in this strategy. As the US grapples with its response, the battle for AI dominance will be a defining feature of US-China relations in the years ahead. The outcome will impact the global technology landscape and influence geopolitical dynamics in an increasingly interconnected world.

Chinese celebrate DeepSeek’s global rise with pride

DeepSeek’s AI assistant has surged ahead of competitors, overtaking ChatGPT in downloads on Apple’s App Store. The Chinese startup’s success has raised concerns among US tech firms while sparking national pride in China. Many see it as a symbol of the country’s resilience against Washington’s tech restrictions.

The company’s ability to offer AI services at a lower cost than OpenAI has prompted discussions about the future of US firms like Nvidia and Microsoft. State media and local commentators have celebrated the achievement, with some arguing it proves US containment efforts have failed. A widely shared article from Zhejiang’s provincial media office argued that China can compete on equal footing in technological innovation.

Public sentiment echoes reactions to Huawei’s surprise 2023 smartphone release, which was seen as defying US sanctions. Social media users have expressed pride in DeepSeek’s rise, with many hoping the country’s AI sector will continue advancing. Some view it as evidence of China’s growing strength in AI.

Startup Waterlily secures funding to expand care planning tech

San Francisco-based startup Waterlily has raised $7 million in seed funding to expand its AI-driven platform for long-term care planning. Founded by Lily Vittayarukskul, the company helps families and financial advisors predict care costs and create tailored financial strategies. Using machine learning and data from government and insurance sources, Waterlily provides personalised recommendations on funding options, such as life insurance and long-term care policies.

Waterlily’s technology was inspired by Vittayarukskul’s personal experience of caring for her aunt, which exposed the financial and emotional strain of long-term care. The platform’s predictive AI can be used for individuals over 40, offering insights into when and how they may need care. The startup already serves major insurance carriers, including Prudential, and hundreds of independent advisors.

With its latest funding round, Waterlily plans to enhance its AI models, expand its team, and strengthen its partnerships. The company is also exploring international expansion to markets such as the UK and Canada, aiming to bridge the gap in long-term care planning and ensure more families are prepared for the future.

DeepSeek: Speeding up the planet or levelling with ChatGPT?

Although the company’s name somewhat overlaps with Google DeepMind, which was launched earlier, the new player in the market has sparked a leap in attention and public interest, becoming one of the biggest AI surprises on the planet upon its launch.

DeepSeek, a company headquartered in China, enjoys significant popularity primarily because its most sought-after features keep pace with giants like OpenAI and Google, as well as due to notable stock market changes that are far from negligible.

In the following points, we will explore these factors and what the future holds for this young company, particularly in the context of the dynamics between China and the US.

How did it start? Origins of DeepSeek

DeepSeek is an AI company from China based in Hangzhou, Zhejiang, founded by entrepreneur and businessman Liang Wenfeng. The company develops open-source LLMs and is owned by a Chinese hedge fund, High-Flyer.

It all started back in 2015 when Liang Wenfeng cofounded High-Flyer. At first, it was a startup, but in 2019, it grew into a hedge fund focused on developing and using AI trading algorithms. For the first two years, they used AI only for trading.

In 2023, High-Flyer founded a startup called DeepSeek, and Liang Wenfeng was appointed CEO. Two years later, on 10 January 2025, DeepSeek announced the release of its first free-to-use chatbot app. The app surpassed its main competitor, ChatGPT, as the most downloaded free app in the US in just 17 days, causing an unprecedented stir on the market.

Unprecedented impact on the market

Few missed the launch of the DeepSeek model, which is why the stock market felt the impact, and so did some of the biggest giants.

For instance, the value of Nvidia shares dropped by as much as 18%. Similar declines were experienced by giants like OpenAI, Google, and other AI companies focused on small and medium-sized enterprises.

On top of this, there is justified concern among investors, who could quickly shift their focus and redirect their investments. However, this could lead to an even more significant drop in the shares of the largest companies.

Open-source approach

DeepSeek embraces an open-source philosophy, making its AI algorithms, models, and training details freely accessible to the public. The company stated that it is committed to transparency and fosters collaboration among developers and researchers worldwide. They also advocate for a more inclusive and innovative AI ecosystem.

Their strategy has the potential to reshape the AI landscape, as it empowers individuals and organisations to contribute to the evolution of AI technology. DeepSeek’s initiative highlights the importance of open collaboration in driving progress and solving complex challenges in the tech industry.

DeepSeek quickly secured the information after being alerted.

With the growing demand for ethical and transparent AI development, DeepSeek’s open-source model sets a precedent for the industry. The company paves the way for a future where AI breakthroughs are driven by collective effort rather than proprietary control.

Cheaper AI model that shook the market

By being cheaper than the competition, DeepSeek has opened the doors of the AI market to many other companies that do not have as much financial power. As dr Jovan Kurbalija, executive director of Diplo, says in his blog post titled ‘How David outwits Goliath in the age of AI?‘, ‘the age of David challenging Goliath has arrived in AI’.

For individuals, this means monthly costs are reduced by 30% to 50%, which can be, and often is, the biggest incentive for users looking to save.

The privileges once enjoyed by those with greater financial resources are now available to those who want to advance their small and medium-sized businesses.

Cyber threats and challenges faced by DeepSeek

Shortly after its launch, DeepSeek faced a significant setback when it was revealed that an error had exposed sensitive information to the public.

This raised alarms for many, especially as the immense popularity led to the AI Assistant being removed from the AppStore more times than OpenAI’s offering, and a large amount of data became accessible.

Experts have expressed concerns that others may have accessed the leaked data. The company has not yet commented on the incident, while the system’s vulnerability provides a foundation for hacking groups to exploit.

DeepSeek for the top spot, ChatGPT defends the throne

The AI race is heating up as DeepSeek challenges industry leader ChatGPT, aiming to claim the top spot in AI. With its open-source approach, DeepSeek is rapidly gaining attention by publicly making its models and training methods available, fostering innovation and collaboration across the AI community.

The race was further spiced up by DeepSeek’s claim that it built an AI model on par with OpenAI’s ChatGPT for under $6 million (£4.8 million). In comparison, Microsoft, OpenAI’s main partner, plans to invest around $80 billion in AI infrastructure this year.

OpenAI’s ChatGPT search tool faces risks of manipulation via hidden content, leading to biased or harmful outputs.

As DeepSeek pushes forward with its transparent and accessible model, the battle for AI supremacy intensifies. Whether openness will outmatch ChatGPT’s established presence remains to be seen, but one thing is sure—the AI landscape is evolving faster than ever.

Why is DeepSeek gaining popularity in 2025?

DeepSeek has emerged as a major player in AI by embracing an open-source philosophy, making its models and training data freely available to developers. This transparency has fueled rapid innovation, allowing researchers and businesses to build upon its technology and contribute to advancements in AI.

Unlike closed systems controlled by major tech giants, DeepSeek’s approach promotes accessibility and collaboration, attracting a growing community of AI enthusiasts. Its cost-effective development, reportedly achieving results comparable to top-tier models with significantly lower investment, has also drawn attention.

As the demand for more open and adaptable AI solutions rises, DeepSeek’s commitment to shared knowledge positions it as a strong contender in the industry. Whether this strategy will redefine the AI landscape remains to be seen, but its growing influence in 2025 is undeniable.

DeepSeek in the future: Development, features, and strategies

Now that it has experienced ‘overnight success,’ the Chinese company aims to push DeepSeek to the top and position it among the most powerful AI firms in the world.

Users can definitely expect many advanced features that will fuel a fierce battle with giants like DeepMind and ChatGPT.

Strategically, DeepSeek will attempt to break into the American market and offer more financially accessible solutions, forcing the key players to make significant cuts.

DeepSeek is undoubtedly a real hit in the market, but it remains to be seen whether price is the only measure of its success.

Whether it will make a leap in its own technology and completely outpace the competition or remain shoulder to shoulder with the giants—or even falter—will be revealed in the near future.

One thing is sure: the Chinese company has seriously shaken up the market, which will need considerable time to recover.

DeepSeek data exposed online before swift removal

A cybersecurity firm Wiz has discovered that Chinese AI startup DeepSeek inadvertently exposed sensitive data online. New York-based Wiz found more than a million lines of unsecured information, including digital software keys and chat logs capturing user interactions with the company’s AI assistant.

DeepSeek acted swiftly to secure the data after Wiz reported the issue. The company’s chief technology officer noted that the exposure was easy to find, raising concerns that others may have accessed the information before it was taken down. DeepSeek has not commented on the incident.

The startup has gained rapid popularity, with its AI assistant surpassing ChatGPT in downloads from Apple’s App Store. Its rise has intensified competition in the AI sector, sparking debates about the sustainability of US tech giants’ business models and profit margins.

Diplo chatbot can make mistakes. Consider checking important information.