After gaining attention with Studio Ghibli-style portraits, ChatGPT’s image generator is now powering a new wave of self-representation online – this time with users turning themselves into plastic action figures.

What began as a quirky trend on LinkedIn has now spread to platforms like Instagram, TikTok, Facebook, and X. The trend includes different takes, but the “AI Action Figure” version is among the most common. It typically shows a person recreated as a doll encased in a plastic blister pack, often accessorised with work-related items like laptops, books, or coffee mugs. That’s fitting, considering the trend’s initial traction among professionals and marketers on LinkedIn.

Other versions draw inspiration from more recognisable aesthetics, like the “Barbie Box Challenge,” where the AI-generated figure is styled to resemble a vintage Barbie.

The rise of the virtual dolls follows the earlier success of the Studio Ghibli-style portraits, which pushed ChatGPT’s image capabilities into the spotlight. That earlier trend sparked some backlash related to environmental, copyright, and creative concerns – but so far, the doll-themed offshoot hasn’t drawn the same level of criticism.

What’s notable about the trends is the consistent use of ChatGPT as the generator of choice. OpenAI’s recent launch of GPT-4o, which includes native image generation, attracted such a large volume of users that the firm had to temporarily limit image output and delay rollout for free-tier accounts.

While the popularity of action figures hasn’t yet matched that of Ghibli portraits, it does highlight ChatGPT’s role in introducing image tools to a broader user base. Many of these doll images are shared by users with low engagement, and mostly in professional circles. Some brands, including Mac Cosmetics and NYX, have posted their own versions, but celebrities and influencers have largely stayed away. One notable exception is US Representative Marjorie Taylor Greene, who shared a version of herself with accessories including a Bible and a gavel, calling it “The Congresswoman MTG Starter Kit.”

What the AI Barbie trend looks like

The process involves uploading a photo into ChatGPT and prompting it to create a doll or action figure based on the image. Many users opt for the Barbie aesthetic, asking for stylised packaging and accessories that reflect their personal or professional identity. The final output often mimics retro Barbie ads from the 1990s or early 2000s. Participants typically specify details like:

• The name to be displayed on the box
• Accessories, like pets, smartphones, or coffee mugs
• The desired pose, facial expression, or outfit
• Packaging design elements like colour or slogans

Users often iterate through several versions, adjusting prompts to better match their expectations. The theme can vary widely – from professional personas to hobbies or fictional characters – giving the trend a broad creative range.

How the trend gained momentum

The idea gained visibility in early 2025, beginning on LinkedIn where users embraced the “AI Action Figure” format. The Barbie-style makeover gained traction over time, tapping into a blend of nostalgia and visual novelty. Hashtags like #aibarbie and #BarbieBoxChallenge have helped to spread the concept. While the Barbie-inspired version has not gone as viral as the Ghibli-style portraits, it has maintained steady traction on social media, especially among users looking for lighthearted ways to express their personal branding.

https://youtube.com/watch?v=Z6S6zQQ8sCQ%3Fsi%3DPJOwLgHWngf21YhL

Using ChatGPT’s image tool

To participate, users must access ChatGPT’s image generation tool, available with GPT-4o. The process begins by uploading a high-resolution photo – preferably full-body – and supplying a prompt that describes the desired figurine.

To improve accuracy, prompts usually include:

• A theme (e.g., office, workout, fantasy)
• Instructions for how the figure should be posed
• Details about clothing, mood, or accessories
• A note to include these elements inside a moulded box layout

Reiterating the intended theme helps ensure consistent results. While many focus on work-related personas, the style is flexible – some choose gym-themed versions, others opt for more humorous or fictional spins.

Behind the spike in image activity

ChatGPT’s image generation tool launched widely in early 2025, and its use quickly surged. According to OpenAI CEO Sam Altman, the demand became so intense that GPU capacity was stretched thin, prompting a temporary cap on image generation for free users. Altman described the load as “biblical demand” in a social media post, noting that the feature had drawn more than 150 million active users in its first month. The tool’s ability to generate everything from cartoons to logos – and now custom action figures – has played a central role in how users explore visual identity through AI.

The post First it was Ghibli, now it’s the AI Barbie Box trend appeared first on TechWire Asia.

The artificial intelligence race between China and the United States has entered a new phase as Chinese companies narrow the technology gap despite Western sanctions.

According to Lee Kai-fu, CEO of Chinese startup 01.AI and former head of Google China, the gap in core technologies has shrunk from “six to nine months” to “probably three months,” with China actually pulling ahead in specific areas like infrastructure software engineering. The Chinese AI startup DeepSeek has become the epicentre of the intensifying technological rivalry.

On January 20, 2025, while the world’s attention was fixed on Donald Trump’s inauguration, DeepSeek quietly launched its R1 model – a low-cost, open-source, high-performance large language model with capabilities reportedly rivalling or surpassing OpenAI’s ChatGPT-4, but at a fraction of the cost.

“The fact that DeepSeek can figure out the chain of thought with a new way to do reinforcement learning is either catching up with the US, learning quickly, or maybe even more innovative now,” Lee told Reuters, referring to how DeepSeek models show users their reasoning process before delivering answers.

Innovative efficiency: China’s response to chip sanctions

DeepSeek’s achievement is particularly notable because it emerged despite US restrictions on advanced processor chip exports to China. Instead of being hampered by international limitations, Chinese companies have responded by optimising efficiency and compensating for lower-quality hardware with quantity.

The adaptive approach was demonstrated further on March 25, 2025, when DeepSeek upgraded its V3 large language model. The new version, DeepSeek-V3-0324, features enhanced reasoning capabilities, optimised front-end web development, and upgraded Chinese writing proficiency. DeepSeek-V3-0324 significantly improved in several benchmark tests, especially in mathematics and coding. Häme University lecturer Kuittinen Petri highlighted the significance of these advancements, stating on social media:

“DeepSeek is doing all this with just [roughly] 2% [of the] money resources of OpenAI.” He added that when he asked the new model to “create a great-looking responsive front page for an AI company,” it produced a mobile-friendly, properly functioning website after coding 958 lines.

Global market implications

The impact of China’s AI advances extends beyond technological achievement to financial markets. When DeepSeek launched its R1 model in January, America’s Nasdaq plunged 3.1%, while the S&P 500 fell 1.5%, demonstrating the wider economic significance of technological competition.

The AI race presents opportunities and challenges for Asia and other regions. China’s low-cost, open-source model could help emerging economies develop AI innovation and entrepreneurship. It also pressures closed-source firms like OpenAI to reconsider their stance.

Meanwhile, both superpowers are making massive investments in AI infrastructure. The Trump administration has unveiled the $500 billion Stargate Project, and China is projected to invest more than 10 trillion yuan (US$1.4 trillion) into technology by 2030.

A double-edged sword for global technology

The US-China tech rivalry risks deepening global divides, forcing nations to navigate growing complexities. Countries face difficult questions: How can they manage research partnerships with China without jeopardising collaboration with US institutions?

How can nations reliant on Chinese materials and exports avoid Chinese technologies? South Korea, the world’s second-largest producer of semiconductors, labours with this dilemma. In 2023, it became more dependent on China for five of the six important raw materials needed for chip-making. Major firms like Toyota, SK Hynix, Samsung, and LG Chem remain vulnerable due to Chinese supply chain dominance. And, the climate implications of this AI race are significant.

According to the Institute for Progress, maintaining AI leadership will require the United States to build five-gigawatt clusters in the next five years. By 2030, data centres could consume 10% of US electricity, more than double the 4% recorded in 2023.

The path forward

As the AI landscape evolves, DeepSeek’s arrival has challenged the assumption that US sanctions were constraining China’s AI sector. Washington’s semiconductor sanctions have proven to be what Lee Kai-fu calls a “double-edged sword” that created short-term challenges and forced Chinese firms to innovate under constraints.

The rapid development of Chinese AI has reignited debates over US chip export controls. Critics argue that the present restrictions have accelerated China’s domestic innovation, as evidenced by DeepSeek’s development and improving capabilities.

China is demonstrating remarkable resilience and innovation in the face of restrictions. As DeepSeek prepares to launch its R2 model potentially early, the technology gap continues to narrow.

The post AI race intensifies: China narrows the gap appeared first on TechWire Asia.

OpenAI and Google are pushing the US government to allow AI models to train on copyrighted material, arguing that ‘fair use’ is critical for maintaining the country’s competitive edge in artificial intelligence.

Both companies outlined their positions in proposals submitted this week in response to a request from the White House for input on President Donald Trump’s “AI Action Plan.”

The post OpenAI and Google seek approval to train AI on content without permission appeared first on TechWire Asia.

Chinese AI agent Manus has emerged as the latest breakthrough in artificial intelligence technology, generating a significant buzz in the global tech community for its advanced autonomous capabilities.

Developed by Tencent Holdings-backed startup Butterfly Effect, Manus claims to be the world’s first general AI agent, distinguishing itself from conventional chatbots by focusing on task completion rather than just conversation. Unlike traditional AI assistants that primarily handle dialogue, the Chinese AI agent Manus utilises multiple AI models to perform complex, multi-step tasks autonomously.

During its invitation-only online preview last week, Manus demonstrated impressive capabilities, including creating customised websites, screening résumés, and producing property recommendation reports based on specific criteria. Manus executes complex tasks from start to finish, delivering complete solutions rather than just suggestions.

“The video says the AI agent is more advanced than a chatbot because it doesn’t only generate ideas but delivers tangible results, like producing a report recommending properties to buy based on specific criteria,” reported CNN Business.

How Manus works

In a post on X, Peak Ji Yichao, co-founder and chief scientist at Butterfly Effect, stated the Chinese AI agent Manus was built using existing large language models, including Anthropic’s Claude and fine-tuned versions of Alibaba’s open-source Qwen. The multi-model approach enables more sophisticated reasoning and action capabilities.

According to MIT Technology Review, which obtained access to the platform, using Manus “feels like collaborating with a highly intelligent and efficient intern.”

The publication tested Manus on three tasks: compiling a list of journalists covering China tech, searching for property listings with specific criteria, and nominating candidates for MIT Technology Review’s Innovators Under 35 list.

“While it occasionally lacks understanding of what it’s being asked to do, makes incorrect assumptions, or cuts corners to expedite tasks, it explains its reasoning clearly, is remarkably adaptable, and can improve substantially when provided with detailed instructions or feedback,” wrote Caiwei Chen in MIT Technology Review. What differentiates Manus is “Manus’s Computer” window, which allows users to observe what the agent is doing and intervene at any point. The transparency and collaborative machine-human approach gives users greater control over the AI’s actions.

Limited access despite growing interest

Despite generating significant buzz, access to Manus remains highly restricted. Fewer than 1% of users on the wait-list have received an invite code. The platform’s Discord channel has attracted more than 186,000 members, indicating substantial interest from potential users worldwide.

According to Chinese media outlet 36Kr, Manus’s per-task cost is approximately $2, which is claimed to be one-tenth of the cost of OpenAI’s DeepResearch service. If its server infrastructure improves, this cost efficiency could position the Chinese AI agent Manus as a preferred choice for individual professionals and small teams.

The system isn’t without limitations, however. MIT Technology Review reported that Manus encountered obstacles with pay-walled content and experienced occasional crashes due to high service loads. The message “Due to the current high service load, tasks cannot be created. Please try again in a few minutes” appeared multiple times during testing.

Strategic partnership with Alibaba

Recognising the potential for growth in the Chinese market, Manus recently announced a strategic partnership with Alibaba. According to the South China Morning Post, the March 11 announcement revealed “Manus will engage in strategic cooperation with Alibaba’s Qwen team to meet the needs of Chinese users. ”

The two companies are working closely to ensure all Manus functions are available on “domestic models and computing platforms,” although no specific launch date was provided. A representative of Alibaba’s cloud computing unit confirmed the cooperation, saying both parties are collaborating on open-source models and looking forward to “working with more global AI innovators.”

Alibaba’s QwQ-32B reasoning model advances in parallel

The partnership comes when Alibaba is progressing AI development. On March 6, the company unveiled its latest reasoning model, QwQ-32B, claiming its capabilities surpass OpenAI’s o1-mini and rival DeepSeek’s R1 model. The news was impactful enough to boost Alibaba’s Hong Kong-listed shares by 8% on the announcement day.

“Alibaba touted its new model, QwQ-32B, in an online statement as delivering exceptional performance, almost entirely surpassing OpenAI-o1-mini and rivalling the strongest open-source reasoning model, DeepSeek-R1,” reported CNN Business.

What makes this development notable is the model’s efficiency – Alibaba claims QwQ-32B contains just 32 billion parameters compared to DeepSeek R1’s 671 billion, suggesting significantly lower computational requirements. The tech giant intends to invest at least 380 billion yuan ($52.4 billion) in AI and cloud computing infrastructure over the next three years, underscoring its commitment to establishing leadership in AI.

China’s AI ecosystem gains momentum

As the Chinese AI agent Manus continues to develop through its partnership with Alibaba, it symbolises a broader phenomenon. The Chinese government’s policies include promoting technological self-reliance, generous funding initiatives, and a robust pipeline of AI graduates from top Chinese universities, which together have created a fertile ecosystem for progression.

Chinese firms are now creating increasingly advanced large language models that compete directly with Western counterparts. The Chinese government’s pledge to support “emerging industries and industries of the future” with increased funding for artificial intelligence, combined with Alibaba’s massive $52.4 billion investment commitment, demonstrates the nation’s strategic prioritisation of AI development.

The centrally-coordinated approach is yielding results. The emergence of tools like Manus suggests that the global AI landscape is evolving into a genuinely multi-polar space, with Chinese innovations increasingly setting trends rather than following them.

The post Chinese AI agent Manus is breakthrough in autonomous AI technology appeared first on TechWire Asia.

According to a person familiar with the matter, Microsoft is working on in-house AI models that could compete with those from industry leaders like its partner, OpenAI.
Microsoft has tested a family of models it calls MAI, which reportedly produced results on a par with state-of-the-art AI models from OpenAI and Anthropic. Redmond is looking at how these models might support products like its Copilot-branded AI assistants, which handle user queries and provide suggestions for tasks like document editing and conference calls.

In addition to MAI, Microsoft is working on reasoning models designed to tackle complex problems and simulate human-like decision-making. OpenAI, Anthropic, and Alphabet are also developing similar models.

Microsoft incorporated OpenAI’s o1 reasoning model into its Copilot products last month. A Microsoft spokesperson stated that the company continues to use a mix of models from OpenAI, Microsoft AI, and open-source sources to support its products.

Reducing dependence on OpenAI

The development of MAI models reflects Microsoft’s broader effort to reduce reliance on OpenAI. It has invested around $13 billion in OpenAI since forming a partnership in 2019, which gave OpenAI access to Microsoft’s Azure cloud platform to power its AI research and development. The partnership between the two companies was renegotiated in January, allowing OpenAI to use cloud services from competitors unless Microsoft claims the business for itself. The updated agreement runs until 2030.

Amy Hood, Microsoft’s Chief Financial Officer, recently spoke about the partnership at a Morgan Stanley conference. “We’re both successful when each of us are successful,” Hood said. “So as you go through that process, I do think everybody’s planning for what happens for a decade, or two decades. And that’s important for both of us to do.”

OpenAI’s role in Microsoft’s products

>Since the partnership began, OpenAI’s models have been integrated into Microsoft products, including Office, GitHub Copilot, and Bing Search. Microsoft’s AI infrastructure is primarily hosted on Azure, and the company also collaborates with OpenAI on AI supercomputing and large language models (LLMs). “We feel great about having leading models from OpenAI, we’re still incredibly proud of that,” Hood previously said. “But we also have other models, including ones we build, to make sure that there’s choice.”

Expanding AI model options

Alongside OpenAI’s models, Microsoft has developed a set of smaller in-house models called Phi, and tested AI models from other providers, including Anthropic, DeepSeek, Meta, and Elon Musk’s xAI, to evaluate their performance in the Copilot framework.

Anthropic’s Claude is known for its focus on safety and alignment with human values. The company has secured significant investment, raising its valuation to $61.5 billion. Google’s Gemini, developed by DeepMind, is a multimodal model capable of processing text, images, audio, and video simultaneously. Google has positioned it as a competitor to OpenAI’s GPT-4, with multiple versions like Gemini Ultra, Pro, and Nano tailored to different use cases.

Meta’s LLaMA series is an open-source model designed to foster transparency and accessibility for developers. Meta’s focus has been on creating conversational AI with natural interactions. xAI’s Grok, integrated into the X platform, focuses on real-time information and conversational AI. It is positioned as a direct competitor to ChatGPT and other conversational models.

Microsoft’s decision to develop its own models reflects the growing demand for diversified AI capabilities. By expanding its model portfolio, the company aims to offer more flexibility and reduce its dependence on a single partner. The development of MAI models positions Microsoft alongside other major players in the AI market, increasing its ability to respond to shifting industry demands and technological advances.

A balanced approach to AI development

Microsoft’s strategy to combine in-house models with external solutions highlights its effort to balance independence with strategic partnerships. OpenAI remains an important partner, but the development of MAI models positions Microsoft to adapt to shifts in the AI market and meet increasing demand for more versatile AI solutions.

The post Microsoft develops in-house AI models to compete with OpenAI appeared first on TechWire Asia.

DeepSeek AI in Chinese EVs is quickly becoming the new battleground for automakers, with companies trying to secure their competitive edge in China’s cutthroat electric vehicle market. AI technology has emerged as a important differentiator, and Phate Zhang, founder of CnEVPost, warns that vehicles without DeepSeek integration risk being “edged out of the market” entirely.

The South China Morning Post reports that DeepSeek AI adoption in Chinese EVs has gained momentum, with over a dozen automakers announcing integration plans in recent weeks. The technology has attracted manufacturers, from EV giant BYD to emerging players like the Stellantis-backed Leapmotor.

Integrating DeepSeek AI in Chinese EVs represents a significant technological leap forward. The Hangzhou-based AI company has captured global attention with its recent release of two advanced open-source AI models: DeepSeek-V3 and DeepSeek-R1.

The models are particularly notable for their ability to operate at a fraction of the cost and computing power typically required for large language model (LLM) projects, making them an attractive option for automakers looking to enhance their vehicles’ capabilities.

BYD, China’s leading EV manufacturer, has announced plans to integrate DeepSeek with its Xuanji vehicle software. The company aims to offer preliminary self-driving capabilities in nearly all its models at no additional cost, making autonomous driving features more accessible to mainland Chinese customers.

Its strategy includes equipping at least 21 models – including the affordable Seagull hatchback, priced at 69,800 yuan (US$9,575) – with Indigenous Advanced Driver Assistance Systems (ADAS).

The widespread adoption of DeepSeek AI in Chinese EVs extends beyond BYD. Other major players in the Chinese automotive industry, including Geely, Great Wall Motor, Chery Automobile, and SAIC Motor, have also announced plans to incorporate DeepSeek’s AI into their digital cockpits, highlighting the industry-wide adoption of this technology.

According to research firm IDC, DeepSeek’s open-source model has fostered a collaborative innovation ecosystem through platforms like GitHub, where global developers participate in optimisation and security testing.

The collaborative approach is expected to significantly improve the ability to effectively deploy, train, and utilise large language models. DeepSeek AI’s market impact on Chinese EVs is expected to be substantial. Zhang Yongwei, general secretary of China EV100, projects that by 2025, approximately 15 million cars – representing two-thirds of national sales – will be equipped with preliminary autonomous driving systems.

The forecast underscores the rapid transformation of China’s automotive landscape through AI integration. The competitive pressure in China’s EV market intensifies, as evidenced by aggressive pricing strategies. The China Passenger Car Association reports that manufacturers reduced prices on 195 models between January and November last year, compared to 150 models in 2023.

In this context, DeepSeek AI integration has become a important differentiator for automakers seeking to maintain their market position. As the integration of DeepSeek AI in Chinese EVs accelerates, the technology appears to be not just a trend but a necessary evolution.

The movement potentially sets new standards for the global automotive industry.

The post Why are Chinese EV manufacturers racing to integrate DeepSeek’s AI technology? appeared first on TechWire Asia.

Chinese iPhone users may soon experience AI features powered by home-grown technology, as Apple Inc. is reportedly planning to integrate Alibaba’s Qwen AI models into its devices. The move comes as the tech giant seeks to bridge a critical gap in its Chinese market offerings, where its global Apple Intelligence features remain unavailable due to regulatory restrictions.

The potential partnership, first reported by The Information and subsequently detailed by the South China Morning Post, suggests that Apple has chosen Alibaba’s Qwen AI model to power its AI features in China. This decision reflects Apple’s pragmatic approach to maintaining its competitive edge in the world’s largest smartphone market, while also adhering to local regulations.

Strategic implications for both giants

Industry analysts view this potential collaboration as a win-win situation for both companies. For Apple, partnering with a local AI powerhouse could help navigate China’s complex regulatory landscape while ensuring its devices remain competitive in a market increasingly dominated by AI-powered smartphones.

For Alibaba, securing Apple as a flagship partner could significantly boost its credibility in the global AI race. “The partnership between Alibaba and Apple partially validates Alibaba’s capabilities in AI and Qwen 2.5 Max,” notes Morningstar’s senior equity analyst Chelsey Lam. 

“Selecting the right partner to deliver a good Apple Intelligence experience will help them reinvigorate iPhone sales in China.”

Technical Prowess and Market Timing

Alibaba’s Qwen AI model has recently gained significant recognition in the global AI community. According to the collaborative machine-learning platform Hugging Face, Qwen is powering the world’s top 10 open-source large language models (LLMs). This technical achievement likely played a crucial role in Apple’s decision to partner with Alibaba.

This development’s timing is particularly noteworthy, as Apple prepares to host its China developer conference in Shanghai on March 25. The event has sparked speculation about the possible launch of Apple Intelligence features for Chinese iPhone users, potentially powered by Alibaba’s Qwen model.

Market response and future implications

The market has responded positively to the news, with Alibaba’s stock price gaining 1.3% in the US on Tuesday, following a 7.6% increase on Monday. Apple’s stock also saw a 2.2% increase, indicating investor confidence in the potential partnership.

The collaboration could have far-reaching implications for both companies. For Apple, it represents a practical solution to maintain its premium positioning in the Chinese market while complying with local regulations. 

As industry experts indicate, this partnership could increase Alibaba’s interest from other global technology companies seeking AI partnerships in China.

Looking ahead

While neither Apple nor Alibaba has officially confirmed the partnership, the development has generated significant buzz among Chinese iOS developers. The potential integration of Qwen AI models into iPhones could open new opportunities for local developers to create AI-powered applications tailored to the Chinese market.

As the March 25 developer conference approaches, the tech community will watch closely for any official announcements regarding this partnership. If confirmed, this collaboration could set a precedent for how international technology companies adapt their AI strategies to succeed in the Chinese market while complying with local regulations.

The reported partnership underscores a growing trend of localisation in AI technology. Global companies are increasingly seeking local partners to maintain their competitive edge in specific markets.

For Apple, whose success in China has been crucial to its global growth story, this adaptation could prove essential in maintaining its market position amid rising competition from local smartphone manufacturers with integrated AI capabilities.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Will Apple Intelligence launch in China through Alibaba’s Qwen AI models? appeared first on TechWire Asia.

OpenAI is shaking things up with a big update to its Model Spec—the rulebook that guides how its AI models should behave. And this time, they’re not keeping it to themselves. The 63-page document is now free for anyone to use, adapt, or build on, giving the broader AI community a peek into how OpenAI thinks about AI behaviour.

This new version is a major expansion from the previous 10-page spec, covering everything from handling controversial topics to giving users more control over how AI interacts with them. The goal? To make AI models more flexible, transparent, and better equipped to let users explore ideas freely—without hitting arbitrary walls.

Why now?

The timing is no coincidence. OpenAI’s CEO, Sam Altman, recently hinted that GPT-4.5 (internally codenamed Orion) is on the way, with GPT-5 not far behind. As AI capabilities grow, so does the pressure to get the rules right. OpenAI’s behaviour team, led by Joanne Jang, says they wanted to get ahead of the curve and address tricky ethical questions that have sparked debates over the past year.

One example? That infamous question about misgendering Caitlyn Jenner to stop a nuclear disaster—a bizarre but revealing prompt that highlighted how hard it can be to program AI to navigate ethics. OpenAI says it’s been rethinking how models should approach these kinds of moral dilemmas.

“We can’t create one model with the exact same set of behaviour standards that everyone in the world will love,” according to Jang in an interview with The Verge. She highlighted that while OpenAI keeps key safety measures in place, users and developers still have flexibility to adjust many aspects of the model’s behaviour.

More flexibility, not fewer safeguards

The updated Model Spec reflects that balance. Users and developers will have more options to tweak AI behaviour—whether that means making the model more formal, casual, or tailored to their specific needs. But OpenAI is clear that some limits will stay. The model won’t encourage self-harm, create deepfakes, or churn out copyrighted content (especially with The New York Times suing OpenAI over content scraping).

Another notable shift is how the company handles controversial topics. Instead of dodging tough questions, the spec encourages models to “seek the truth together” with users—offering thoughtful, reasoned answers while standing firm against misinformation or harmful content.

Mature content and “grown-up mode”

There’s also a more nuanced approach to adult content. After feedback from users asking for a “grown-up mode” (something Altman publicly supported last year), OpenAI is exploring ways to allow certain mature content—like erotica—in appropriate contexts, while keeping strict bans on things like revenge porn or exploitative material.

This is a departure from OpenAI’s previous blanket ban on anything explicit, and it signals that the company is trying to balance creativity and safety.

Fixing “AI sycophancy”

Another area getting attention is AI sycophancy—a fancy way of saying that AI models sometimes agree too easily, even when they shouldn’t. OpenAI wants ChatGPT to feel more like a thoughtful colleague, not a people-pleaser.

That means it should correct users when they’re wrong, offer honest feedback instead of empty praise, and give consistent answers no matter how a question is phrased. The goal is to make interactions more reliable, so users don’t have to game the system to get accurate information. “We don’t ever want users to feel like they have to somehow carefully engineer their prompt to not get the model to just agree with you,” Jang explained.

Who’s in charge?

OpenAI also clarified the chain of command for AI instructions. Platform-level rules come first, followed by developer guidelines, and then individual user preferences. This hierarchy is meant to clear up confusion over what can and can’t be customised when using OpenAI’s models.

Open to the public

Crucially, OpenAI is releasing the entire Model Spec under a Creative Commons Zero (CC0) license, meaning other companies can adopt or adapt it however they like. The company hopes this transparency will spark more industry-wide conversations about AI behaviour—and get feedback from the public.

“We knew that it would be spicy,” Jang admitted. “But I think we respect the public’s ability to actually digest these spicy things and process it with us.”

The company is also open-sourcing the prompts it uses to test whether models are following the guidelines.

What’s next?

While the updated spec doesn’t immediately change how ChatGPT works, it signals where OpenAI is headed—especially with GPT-4.5 and GPT-5 in the pipeline. Altman recently shared on X that GPT-4.5 will be the “last non-chain-of-thought model,” hinting that GPT-5 will be more capable of reasoning through complex problems.

He also promised a simpler product experience, acknowledging that users are frustrated with having to pick between different models. “We hate the model picker as much as you do,” Altman wrote, suggesting OpenAI is moving toward a more unified system.

As all this unfolds, the broader AI industry will be watching closely. With Elon Musk trying (and failing) to buy OpenAI’s nonprofit arm for nearly $100 billion, and legal battles mounting, the pressure is on. But for now, OpenAI is betting that openness and user input will help shape the future of AI behaviour—for the better.

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post OpenAI updates AI behaviour guidelines while preparing for GPT-5 appeared first on TechWire Asia.

The AI investment surge among tech giants shows no signs of slowing. Amazon, Microsoft, Google, and Meta still appear to be planning investment of over $320 billion in AI infrastructure by 2025, even as Chinese startup DeepSeek demonstrates potentially more efficient AI model development methods.

The unprecedented level of capital expenditure, up from $246 billion in 2024, reflects Big Tech’s conviction that AI represents a transformative opportunity, one that justifies massive infrastructure investment. The spending plans persist despite investor concerns about DeepSeek’s recent claims that it can train and operate significantly more efficient AI models.

Amazon is leading the investment. The company plans to allocate over $100 billion in capital expenditures this year, marking a significant increase from $77 billion in 2024. The company’s CEO, Andy Jassy, defended the aggressive spending strategy during a recent earnings call, characterising AI as a “once-in-a-lifetime type of business opportunity” for Amazon Web Services (AWS).

Microsoft is also positioning itself as a major contender in the AI cloud race, and has earmarked $80 billion for AI-related infrastructure in 2025. CEO Satya Nadella points to compelling evidence supporting the investment strategy, noting that Microsoft’s AI business “has surpassed an annual revenue run rate of $13 billion, up 175% year-over-year.”

Nadella’s perspective on AI demand draws from the Jevons paradox – the economic observation that increased efficiency in resource use leads to higher, not lower, consumption. “As AI gets more efficient and accessible, we will see its use skyrocket, turning it into a commodity we just can’t get enough of,” Nadella stated in a recent social media post.

Google’s parent company, Alphabet, has also adopted on aggressive stance, planning approximately $75 billion in capital expenditures for 2025, significantly exceeding analysts’ expectations of $58 billion. Despite concerns about the company’s AI strategy and slowing cloud growth, CEO Sundar Pichai maintains that Google is “building, testing, and launching [AI] products faster than ever.”

Meta’s approach to the AI investment surge stands out. The company plans to spend $60-65 billion on capital in 2025, up from $39 billion in 2024. The company’s strategy of pursuing an “American standard” for open-source AI models has resonated with investors, particularly given Meta’s demonstrated ability to monetise AI through advanced ad targeting.

Industry analysts are divided over the impact of DeepSeek’s developments on Big Tech’s AI investments. Jesse Cohen, senior analyst at Investing.com, voiced investor concerns about the need for “clearer timelines on when AI spending translates to earnings and sales growth, not just promises.”

However, Dan Ives, managing director at Wedbush Securities, dismisses DeepSeek as a serious threat to Big Tech’s AI ambitions. “This is an AI arms race, and the Temu of AI DeepSeek is not changing that… [The] AI Revolution [is] just starting.”

The persistence and continued growth of AI investment suggests that significant technology companies view DeepSeek’s efficiency gains as a catalyst for broader AI adoption rather than a threat to their infrastructure investments.

The market response to the massive levels of investment has been mixed. While Meta’s proposed investments in AI have been well-received (shares rose following its earnings call), other technology firms have faced more sceptical investors. After announcing its spending plans, Amazon’s stock fell more than 5%, while Alphabet’s share value dropped more than 8% after its earnings report announcement.

Despite the market’s reactions, technology leaders remain convinced that AI investment represents a strategic necessity rather than an optional luxury. As the industry continues to evolve, the race to build AI infrastructure is intensifying rather than slowing, suggesting that DeepSeek’s developments may indeed be accelerating, not dampening the industry’s AI ambitions.

The post AI investments to reach new heights despite the DeepSeek challenge appeared first on TechWire Asia.

A dramatic shift in open-source AI development is emerging from China, where startup DeepSeek fundamentally challenges how artificial intelligence evolves. By making its technology freely available for anyone to download, modify, and build on, DeepSeek has ignited a debate about the future of AI and global technological leadership.

The cost revolution in open-source AI development

When DeepSeek’s R1 model overtook ChatGPT as the most downloaded free app on the US Apple App Store in January 2025, it represented more than just market success – it signalled a potential shift in AI economics.

At just $5.6 million in development costs – roughly one-tenth the price of Meta’s similarly open-source Llama model – DeepSeek has demonstrated that cutting-edge AI doesn’t require massive investments. Unlike established players like OpenAI and Anthropic, whose models remain closely guarded secrets, DeepSeek opted for transparency, offering its code and technical documentation to the global developer community.

The approach to open-source AI development hasn’t compromised capability; the model demonstrates reasoning and mathematical skills comparable to more expensive rivals. The implications extend beyond finances. Traditional AI development requires massive computing resources, contributing to high electricity consumption and carbon emissions.

DeepSeek’s efficient training method could pioneer more sustainable AI scaling, which is particularly important as these technologies become ubiquitous. However, the “open-source” designation demands scrutiny. According to the Open Source Initiative (OSI), true open-source AI must provide detailed information about training data and allow unrestricted study, use, and modification of the system.

While DeepSeek has released its model weights and some technical documentation, it hasn’t fully disclosed its training data, raising questions about complete transparency.

Global impact and security concerns

The democratisation of AI through cost-effective open-source development could accelerate innovation worldwide. Smaller companies and developers in regions with limited access to cutting-edge technology can now build on established models rather than starting from scratch. The collaborative approach could catalyse more diverse applications and solutions to real-world problems.

Yet this accessibility comes with risks. Critics, particularly in the US, worry about potential misuse of open-source AI models, from developing bioweapons to spreading misinformation. The controversy intensified when OpenAI announced an investigation into whether DeepSeek may have “inappropriately distilled” its models – essentially using OpenAI’s outputs to train its competing system.

The geopolitical implications are significant. As Meta CEO Mark Zuckerberg noted, “The is a huge geopolitical competition, and China’s running at it super hard.” DeepSeek’s success has prompted reflection on the traditional American model of proprietary AI development, with some arguing that the US should embrace more open methodologies to maintain technological leadership.

Looking ahead, DeepSeek’s breakthrough, with its $5.6 million price tag, could reshape the entire AI landscape. The company has proven that efficient, cost-effective open-source AI development is possible without the massive resources typically associated with cutting-edge models.

This could lead to a more diverse and competitive AI ecosystem, with innovation emerging from unexpected quarters. However, the path forward remains complex. Regulatory frameworks are evolving as governments struggle to balance innovation with security concerns worldwide.

The Trump administration has yet to fully articulate its AI policy, though some officials have supported open-source development. What’s clear is that DeepSeek’s efficient approach to open-source AI development has challenged fundamental assumptions about the field, forcing the industry to reconsider what’s possible with limited resources and an open approach.\

Looking to revamp your digital transformation strategy?

Learn more about Digital Transformation Week taking place in Amsterdam, California, and London. The comprehensive event is co-located with IoT Tech Expo, AI & Big Data Expo, Cyber Security & Cloud Expo, and other leading events.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Is open-source AI the future? DeepSeek’s $5.6M challenge to Big Tech appeared first on TechWire Asia.