OpenAI

ChatGPT now supports MCP for business data access, but safety risks remain

OpenAI has officially enabled support for Anthropic’s Model Context Protocol (MCP) in ChatGPT, allowing businesses to connect their internal tools directly to the chatbot through Deep Research.

The development enables employees to retrieve company data from previously siloed systems, offering real-time access to documents and search results via custom-built MCP servers.

Adopting MCP — an open industry protocol recently embraced by OpenAI, Google and Microsoft — opens new possibilities and presents security risks.

OpenAI advises users to avoid third-party MCP servers unless hosted by the official service provider, warning that unverified connections may carry prompt injections or hidden malicious directives. Users are urged to report suspicious activity and avoid exposing sensitive data during integration.

To connect tools, developers must set up an MCP server and create a tailored connector within ChatGPT, complete with detailed instructions. The feature is now live for ChatGPT Enterprise, Team and Edu users, who can share the connector across their workspace as a trusted data source.

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

Meta offers $100M bonuses to poach OpenAI talent but Altman defends mission-driven culture

Meta has reportedly attempted to lure top talent from OpenAI with signing bonuses exceeding $100 million, according to OpenAI’s CEO Sam Altman.

Speaking on a podcast hosted by his brother, Jack Altman, he revealed that Meta has offered extremely high compensation to key OpenAI staff, yet none have accepted the offers.

Meta CEO Mark Zuckerberg is said to be directly involved in recruiting for a new ‘superintelligence’ team as part of the latest AI push.

The tech giant recently announced a $14.3 billion investment in Scale AI and brought Scale’s CEO, Alexandr Wang, on board. Altman believes Meta sees ChatGPT not only as competition for Google but as a potential rival to Facebook regarding user attention.

Altman questioned whether such high-compensation strategies foster the right environment, suggesting that culture cannot be built on upfront financial incentives alone.

He stressed that OpenAI prefers aligning rewards with its mission instead of offering massive pay packets. In his view, sustainable innovation stems from purpose, not payouts.

While recognising Meta’s persistence in the AI race, Altman suggested that the company will likely try again if the current effort fails. He highlighted a cultural difference, saying OpenAI has built a team focused on consistent innovation — something he believes Meta still struggles to understand.

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

OpenAI brings DALL-E image creation to WhatsApp users worldwide

OpenAI has officially launched image creation capabilities for WhatsApp users, expanding access to its AI visual tools via the verified number +1-800-ChatGPT. Using natural language prompts, the feature enables users to generate or edit images directly within their chats.

Previously limited to the web and mobile versions of ChatGPT, the image generation tool—powered by DALL-E—is now available globally on WhatsApp, free of charge. OpenAI announced the rollout via X, encouraging users to connect their accounts for enhanced functionality.

To get started, users should save +1-800-ChatGPT (+1-800-242-8478) to their contacts, send ‘Hi’ via WhatsApp, and follow the instructions to link their OpenAI account.

Once verified, they can prompt the AI with creative requests such as ‘design a futuristic skyline’ or ‘show a dog surfing on Mars’ and receive bespoke visuals in return.

The move further integrates generative AI into everyday messaging, making powerful image-creation tools more accessible to a broad user base.

Meanwhile, WhatsApp is preparing to introduce in-app advertising. With over two billion active users, Meta plans to monetise the platform more aggressively—signalling a notable shift in WhatsApp’s strategy.

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

ChatGPT loses chess match to Atari 2600

ChatGPT, was trounced in a chess match by a 1979 video game running on an Atari 2600 emulator. Citrix engineer Robert Caruso set up the match using Video Chess to test how the AI would perform against vintage gaming software.

The result was unexpectedly lopsided. ChatGPT confused rooks for bishops, forgot piece positions and made repeated beginner mistakes, eventually asking for the match to be restarted. Even when standard chess notation was used, its performance failed to improve.

Caruso described the 90-minute session as full of basic blunders, saying the AI would have been laughed out of a primary school chess club. His post highlighted the limitations of ChatGPT’s architecture, which is built for language understanding, not strategic board gameplay.

While the experiment doesn’t mean ChatGPT is entirely useless at chess, it suggests users are better off discussing the game with the bot than challenging it. OpenAI has not yet responded to the light-hearted but telling critique.

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

OpenAI considers antitrust action against Microsoft over AI hosting control

OpenAI reportedly tries to reduce Microsoft’s exclusive control over hosting its AI models, signalling growing friction between the two companies.

According to the Wall Street Journal, OpenAI leadership has considered filing an antitrust complaint against Microsoft, alleging anti-competitive behaviour in their ongoing collaboration. The move could trigger federal regulatory scrutiny.

The tension comes amid ongoing talks over OpenAI’s corporate restructuring. A report by The Information suggests that OpenAI is negotiating to grant Microsoft a 33% stake in its reorganized for-profit unit. In exchange, Microsoft would give up rights to future profits.

OpenAI also wants to revise its existing contract with Microsoft, particularly clauses that grant exclusive Azure hosting rights. The company reportedly aims to exclude its planned $3 billion acquisition of AI startup Windsurf from the agreement, which otherwise gives Microsoft access to OpenAI’s intellectual property.

This developing rift could reshape one of the most influential alliances in AI. Microsoft has invested heavily in OpenAI since 2019 and integrates its models into Microsoft 365 Copilot and Azure services. However, both firms are diversifying.

OpenAI is turning to Google Cloud and Oracle for additional computing power, while Microsoft has begun integrating alternative AI models into its products.

Industry experts warn that regulatory scrutiny or contract changes could impact enterprise customers relying on tightly integrated AI solutions, particularly in sectors like healthcare and finance. Companies may face service disruptions, higher costs, or compatibility challenges if major players shift strategy or infrastructure.

Analysts suggest that the era of single-model reliance may be ending. As innovation from rivals like DeepSeek accelerates, enterprises and cloud providers are moving toward multi-model support, aiming for modular, scalable, and use-case-specific AI deployments.

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

OpenAI turns to Google Cloud in shift from solo AI race

OpenAI has entered into an unexpected partnership with Google, using Google Cloud to support its growing AI infrastructure needs.

Despite being fierce competitors in AI, the two tech giants recognise that long-term success may require collaboration instead of isolation.

As the demand for high-performance hardware soars, traditional rivals join forces to keep pace. OpenAI, previously backed heavily by Microsoft, now draws from Google’s vast cloud resources, hinting at a changing attitude in the AI race.

Rather than going it alone, firms may benefit more by leveraging each other’s strengths to accelerate development.

Google CEO Sundar Pichai, speaking on a podcast, suggested there is room for multiple winners in the AI sector. He even noted that a major competitor had ‘invited me to a dance’, underscoring a new phase of pragmatic cooperation.

While Google still faces threats to its search dominance from tools like ChatGPT, business incentives may override rivalry.

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

Deepseek searches soar after ChatGPT outage

ChatGPT users faced widespread disruption on 10 June 2025 after a global outage hit OpenAI’s services, affecting both the chatbot and associated APIs. OpenAI has yet to confirm the cause, stating only that users are experiencing high error rates and delays.

The blackout halted work for many creative teams who rely on the tool to generate content and meet deadlines. While some were stalled, others turned to alternatives, sparking a surge in interest in rival AI chatbots.

Searches for DeepSeek, a Chinese-developed AI model, jumped 109% to over 2.1 million on the outage day. Claude AI saw a 95% increase in queries, while interest in Google Gemini and Microsoft Copilot also spiked significantly.

Industry experts say the incident underscores the risk of overdependence on a single platform and highlights the growing maturity of competing AI tools. While frustrating for many, the disruption appears to be fuelling broader competition and diversification in the generative AI market.

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

Sam Altman says GPT-4o demand overwhelmed OpenAI’s GPU supply

OpenAI faced a significant infrastructure strain after its GPT-4o image generator went viral for producing Ghibli-style memes. The sudden influx of user demand added a million new users in under an hour, putting immense pressure on the company’s systems.

CEO Sam Altman admitted that OpenAI had to slow feature rollouts and borrow computing power from its research division to keep the service running. The platform temporarily introduced rate limits as it coped with overloaded GPUs.

Altman described the situation as unprecedented, saying no other company has had to manage such intense viral spikes. He noted that image generation with GPT-4o requires significant compute resources, which the company could not fully meet with its current GPU inventory.

Despite the challenges, Altman maintained that OpenAI is committed to managing high user demand while continuing development. The company is also considering watermarking the AI images created by free users to help manage scale and traceability.

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

OpenAI launches powerful o3-Pro model with steep price cuts

OpenAI has rolled out its most advanced AI model yet, o3-Pro, delivering significant improvements in reasoning and task complexity while introducing steep price reductions.

The model is now available to ChatGPT Pro and Team users, with Enterprise and Education access coming next week. Developers can also access it via OpenAI’s API.

O3-Pro is designed for high-performance use across technology, education, and science sectors. It supports advanced capabilities such as web browsing, code execution, file analysis, and memory retention during conversations.

Despite these upgrades, pricing has been reduced drastically—87% lower than o1-Pro—costing just $20 per million input tokens and $80 per million output tokens. The base o3 model has also seen an 80% price cut.

Evaluators consistently rated o3-Pro higher than previous models for clarity, instruction-following, and accuracy, with standout results in benchmarks like AIME 2024 and GPQA Diamond, where it beat Google’s Gemini 2.5 Pro and Claude 4 Opus, respectively.

Although the model lacks image generation and Canvas support, its reasoning capabilities mark a major step forward in OpenAI’s AI offerings.

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

Sam Altman predicts AI will discover new ideas

In a new blog post titled The Gentle Singularity, OpenAI CEO Sam Altman predicted that AI systems capable of producing ‘novel insights’ may arrive as early as 2026.

While Altman’s essay blends optimism with caution, it subtly signals the company’s next central ambition — creating AI that goes beyond repeating existing knowledge and begins generating original ideas instead of mimicking human reasoning.

Altman’s comments echo a broader industry trend. Researchers are already using OpenAI’s recent o3 and o4-mini models to generate new hypotheses. Competitors like Google, Anthropic and FutureHouse are also shifting their focus towards scientific discovery.

Google’s AlphaEvolve has reportedly devised novel solutions to complex maths problems, while FutureHouse claims to have built AI capable of genuine scientific breakthroughs.

Despite the optimism, experts remain sceptical. Critics argue that AI still struggles to ask meaningful questions, a key ingredient for genuine insight.

Former OpenAI researcher Kenneth Stanley, now leading Lila Sciences, says generating creative hypotheses is a more formidable challenge than agentic behaviour. Whether OpenAI achieves the leap remains uncertain, but Altman’s essay may hint at the company’s next bold step.

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