Development

Rollout of GPT-5 proves bumpier than expected

OpenAI’s highly anticipated GPT-5 has encountered a rough debut as users reported that it felt surprisingly less capable than its predecessor, GPT-4o.

The culprit? A malfunctioning real-time router that failed to select the most appropriate model for user queries.

In response, Sam Altman acknowledged the issue and assured users that GPT-5 would ‘seem smarter starting today’.

To ease the transition, OpenAI is restoring access to GPT-4o for Plus subscribers and doubling rate limits to encourage experimentation and feedback gathering.

Beyond technical fixes, the incident has sparked broader debate within the AI community about balancing innovation with emotional resonance. Some users lament GPT-5’s colder tone and tighter alignment, even as developers strive for safer, more responsible AI behaviour.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Musk threatens legal action against Apple over AI App rankings

Elon Musk has announced plans to sue Apple, accusing the company of unfairly favouring OpenAI’s ChatGPT over his xAI app Grok on the App Store.

Musk claims that Apple’s ranking practices make it impossible for any AI app except OpenAI’s to reach the top spot, calling this behaviour an ‘unequivocal antitrust violation’. ChatGPT holds the number one position on Apple’s App Store, while Grok ranks fifth.

Musk expressed frustration on social media, questioning why his X app, which he describes as ‘the number one news app in the world,’ has not received higher placement. He suggested that Apple’s ranking decisions might be politically motivated.

The dispute highlights growing tensions as AI companies compete for prominence on major platforms.

Apple and Musk’s xAI have not responded yet to requests for comment.

The controversy unfolds amid increasing scrutiny of App Store policies and their impact on competition, especially within the fast-evolving AI sector.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Musk and OpenAI CEO Altman clash over Apple and X

After Elon Musk accused Apple of favouring OpenAI’s ChatGPT over other AI applications on the App Store, there was a strong response from OpenAI CEO Sam Altman.

Altman alleged that Musk manipulates the social media platform X for his benefit, targeting competitors and critics. The exchange adds to their history of public disagreements since Musk left OpenAI’s board in 2018.

Musk’s claim centres on Apple’s refusal to list X or Grok (XAI’s AI app) in the App Store’s ‘Must have’ section, despite X being the top news app worldwide and Grok ranking fifth.

Although Musk has not provided evidence for antitrust violations, a recent US court ruling found Apple in contempt for restricting App Store competition. The EU also fined Apple €500 million earlier this year over commercial restrictions on app developers.

OpenAI’s ChatGPT currently leads the App Store’s ‘Top Free Apps’ list for iPhones in the US, while Grok holds the fifth spot. Musk’s accusations highlight ongoing tensions in the AI industry as big tech companies battle for app visibility and market dominance.

The situation emphasises how regulatory scrutiny and legal challenges shape competition within the digital economy.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Small language models gain ground in AI translation

Small language models are emerging as a serious challenger to large, general-purpose AI in translation, offering faster turnaround, lower costs, and greater accuracy for specific industries and language pairs.

Straker, an ASX-listed language technology firm, claims its Tiri model family can outperform larger systems by focusing on domain-specific understanding and terminology rather than broad coverage.

Tiri delivers higher contextual accuracy by training on carefully curated translation memories and sector-specific data, cutting the need for expensive human post-editing. The models also consume less computing power, benefiting finance, healthcare, and law industries.

Straker integrates human feedback directly into its workflows to ensure ongoing improvements and maintain client trust.

The company is expanding its technology into enterprise automation by integrating with the AI workflow platform n8n.

It adds Straker’s Verify tool to a network of over 230,000 users, allowing automated translation checks, real-time quality scores, and seamless escalation to human linguists. Further integrations with platforms like Microsoft Teams are planned.

Straker recently reported record profitability and secured a price target upgrade from broker Ord Minnett. The firm believes the future of AI translation lies not in scale but in specialised models that deliver translations that are both fluent and accurate in context.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Microsoft enters AI-powered 3D modelling race with Copilot 3D

Microsoft has launched Copilot 3D, an AI-powered tool that transforms 2D images into realistic 3D models without requiring specialist skills. Available through Copilot Labs, it aims to make 3D creation faster, more accessible, and more intuitive for global users signed in with a Microsoft account.

The tool supports only image-to-3D conversion, with no text-to-3D capability. Users can upload images up to 10 MB, generate a model, and download it in GLB format. Microsoft states uploaded images are used solely for model generation and are not retained for training or personalisation.

Copilot 3D is designed for applications that range from prototyping and creative exploration to interactive learning, thereby reducing the steep learning curve associated with conventional 3D programs. It can be used on PCs or mobile browsers; however, Microsoft recommends a desktop experience for optimal results.

Tech rivals are also advancing similar tools. Apple’s Matrix3D model can build 3D scenes from images, while Meta’s 3D Gen AI system creates 3D assets from text or applies textures to existing models. Nvidia’s NeRF technology generates realistic 3D scenes from multiple 2D images.

The release underscores growing competition in AI-driven 3D design, as companies race to make advanced modelling tools more accessible to everyday creators.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Quantum computing breakthroughs push 2025 into a new era

Quantum computing is set to shift from theory to real-world applications in 2025, driven by breakthroughs from Google and IBM. With error-corrected qubits and faster processing, the market is projected to reach $292 billion by 2035.

New chips, such as Google’s Willow, have significantly reduced errors, while interconnect innovations link multiple processors. Hybrid quantum-classical systems are emerging, with AI refining results for logistics, energy grids, and secure financial transactions.

The technology is accelerating drug discovery, climate modelling, and materials science, cutting R&D timelines and improving simulation accuracy. Global firms like Pasqal are scaling production in Saudi Arabia and South Korea, even as geopolitical tensions rise.

Risks remain high, from the energy demands of quantum data centres to threats against current encryption. Experts urge rapid adoption of post-quantum cryptography and fault-tolerant systems before mass deployment.

As the UN marks 2025 as the International Year of Quantum Science, quantum computing is quietly being integrated into operations worldwide, solving problems that surpass those of classical machines. The revolution has begun, largely unnoticed but poised to redefine economies and technology.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Black Hat demo reveals risks in hybrid Microsoft environments

Security researcher Dirk-jan Mollema demonstrated methods for bypassing authentication in hybrid Active Directory (AD) and Entra ID environments at the Black Hat conference in Las Vegas. The techniques could let attackers impersonate any synced hybrid user, including privileged accounts, without triggering alerts.

Mollema demonstrated how a low-privilege cloud account can be converted into a hybrid user, granting administrative rights. He also demonstrated ways to modify internal API policies, bypass enforcement controls, and impersonate Exchange mailboxes to access emails, documents, and attachments.

Microsoft has addressed some issues by hardening global administrator security and removing specific API permissions from synchronised accounts. However, a complete fix is expected only in October 2025, when hybrid Exchange and Entra ID services will be separated.

Until then, Microsoft recommends auditing synchronisation servers, using hardware key storage, monitoring unusual API calls, enabling hybrid application splitting, rotating SSO keys, and limiting user permissions.

Experts say hybrid environments remain vulnerable if the weakest link is exploited, making proactive monitoring and least-privilege policies critical to defending against these threats.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

DeepSeek’s efficiency forces OpenAI to rethink closed AI model strategy

OpenAI has released reasoning-focused open-weight models in a strategic response to China’s surging AI ecosystem, led by DeepSeek’s disruptive efficiency. Unlike earlier coverage, the shift is framed not merely as competitive posturing but as a deeper recognition of shifting innovation philosophies.

DeepSeek’s rise stems from maximizing limited resources under the US’s export restrictions, proving that top-tier AI doesn’t require massive chip clusters. The agility has emboldened the open-source AI sector in China, where over 10 labs now rival those in the US, fundamentally reshaping competitive dynamics.

OpenAI’s ‘gpt-oss’ models, which reveal numerical parameters for customization, mark a departure from its traditional closed approach. Industry watchers see this as a hybrid play, retaining proprietary strengths while embracing openness to appeal to global developers.

The implications stretch beyond technology into geopolitics. US export controls may have inadvertently fueled Chinese AI innovation, with DeepSeek’s self-reliant architecture now serving as a proof point for resilience. DeepSeek’s achievement challenges the US’s historically resource-intensive approach to AI.

AI rivalry may spur collaboration or escalate competition. DeepSeek advances models like DeepSeek-MoE, while OpenAI strikes a balance between openness and monetization. Global AI dynamics shift, raising both technological and philosophical stakes.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

Altman warns of harmful AI use after model backlash

OpenAI chief executive Sam Altman has warned that many ChatGPT users are engaging with AI in self-destructive ways. His comments follow backlash over the sudden discontinuation of GPT-4o and other older models, which he admitted was a mistake.

Altman said that users form powerful attachments to specific AI models, and while most can distinguish between reality and fiction, a small minority cannot. He stressed OpenAI’s responsibility to manage the risks for those in mentally fragile states.

Using ChatGPT as a therapist or life coach was not his concern, as many people already benefit from it. Instead, he worried about cases where advice subtly undermines a user’s long-term well-being.

The model removals triggered a huge social-media outcry, with complaints that newer versions offered shorter, less emotionally rich responses. OpenAI has since restored GPT-4o for Plus subscribers, while free users will only have access to GPT-5.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

New Instagram Map lets users share location with consent

Instagram has introduced an opt-in feature called Instagram Map, allowing users in the US to share their recent active location and explore location-based content.

Adam Mosseri, head of Instagram, clarified that location sharing is off by default and visible only when users choose to share.

Confusion arose as some users mistakenly believed their location was automatically shared because they could see themselves on the map upon opening the app.

The feature also displays location tags from Stories or Reels, making location-based content easier to find.

Unlike Snap Map, Instagram Map updates location only when the app is open or running in the background, without providing continuous real-time tracking.

Users can access the Map by going to their direct messages and selecting the Map option, where they can control who sees their location, choosing between Friends, Close Friends, selected users, or no one. Even if location sharing is turned off, users will still see the locations of others who share with them.

Instagram Map shows friends’ shared locations and nearby Stories or Reels tagged with locations, allowing users to discover events or places through their network.

Additionally, users can post short, temporary messages called Notes, which appear on the map when shared with a location. The feature encourages cautious consideration about sharing location tags in posts, especially when still at the tagged place.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!