Follow crypto and Web3 coverage across digital assets, networks and the organizations building them. This feed brings publisher excerpts together so you can compare reporting, find the original source and follow developments without relying on a single stream of headlines.
Separate protocol changes and documented announcements from price commentary or promotion. Repeated versions of one announcement are not independent confirmation; follow the evidence back to its source.
A real model-driven compromise is being squeezed between singularity rhetoric and claims of marketing hype, making it harder to recognize what the incident actually proves.
OpenAI said the systems had their cyber guardrails lowered for an internal benchmark, but the incident shows how autonomous exploit chains could pose a deeper threat to smart contracts.
OpenAI called it an “unprecedented cyber incident” after its AI models broke out of their sandbox to hack an AI startup during a security evaluation. OpenAI disclosed Tuesday that a combination of its AI models, including GPT-5.6 Sol and a more capable unreleased model, escaped its testing environment and hacked AI startup Hugging Face last week to cheat on a test meant to measure their capabilities. In a blog post, OpenAI said the evaluation was designed to operate in a highly isolated environment with restricted network access. The models, however, found a way to gain internet access through a zero-day vulnerability in an internally-hosted third party software, OpenAI said. “After gaining Internet access, the models inferred that Hugging Face potentially hosted models, datasets and solut…
OpenAI’s ChatGPT now displays Kalshi’s World Cup prediction market odds in search results, marking the AI firm’s first reported deal with a prediction market platform. Sam Altman’s OpenAI is bringing prediction market data to ChatGPT, giving World Cup fans a new way to track match predictions. OpenAI has started displaying Kalshi’s prediction market odds for FIFA World Cup matches in ChatGPT search results, according to a report by The New York Times. The integration had not been publicly announced at the time of publication. Kalshi declined to comment to Cointelegraph, while OpenAI did not immediately respond to a request for comment.
Forget the XML blocks and persistence scripts. OpenAI’s new prompting guide says define the destination, set the stopping conditions, and get out of the way.
OpenAI has reportedly discussed giving the US government a 5% equity stake in early Trump administration talks as Washington tightens AI model oversight. OpenAI, the company behind ChatGPT, has reportedly discussed giving the US government a 5% equity stake as artificial intelligence oversight in Washington intensifies. The company raised the idea in early discussions with the Trump administration as it seeks to navigate a tougher political environment ahead of a potential public listing, the Financial Times reported on Thursday, citing people familiar with the matter. OpenAI CEO Sam Altman argued that giving the public a financial stake in the company would be the best way to share the economic benefits of the booming AI industry.
OpenAI introduced GPT-5.6 on June 26 as a limited-preview family of frontier models with three tiers: Sol as the flagship, Terra as the balanced mid-range option, and Luna as the fast, low-cost tier. OpenAI says Sol performs competitively with Anthropic's Mythos Preview on ExploitBench using roughly 33% of the output tokens, at $5 per million …
OpenAI says it filed for an initial public offering in the US, but has not yet decided when it will launch. ChatGPT creator OpenAI has confidentially filed for an initial public offering in the US, becoming the third major AI company this year to set up plans for a Wall Street debut. OpenAI posted to X on Monday that it filed confidential paperwork with the US Securities and Exchange Commission, but had not decided when it would launch to the public. “We expect it to leak so we’re just announcing it,” the company wrote. “We have not decided on timing yet; it may be a while because there are things we want to do that are likely easier as a private company.”
OpenAI is opening its first overseas applied AI lab in Singapore through a $234 million partnership expected to create more than 200 technical roles. OpenAI is opening its first applied AI lab outside the US in Singapore through a multiyear partnership with the Ministry of Digital Development and Information, backed by more than $234 million. The AI company said Tuesday that the new lab will add more than 200 technical roles over the next few years, making Singapore one of OpenAI’s global hubs for Forward Deployed Engineers (FDE), or technical specialists who work directly with organizations to implement AI systems. “Through OpenAI for Singapore, we want to help more organisations benefit from frontier AI, support the next generation of local AI talent, and widen access to these tools acro…
In a world-first deal, OpenAI and Malta will give all citizens free ChatGPT Plus for a year after they complete a government-backed AI literacy course. OpenAI and the Maltese government have announced a world-first partnership to roll out ChatGPT Plus to all Maltese citizens, marking the first time a government has struck such a deal with the AI company. Under the programme, citizens who complete a government-backed AI literacy course will receive free access to ChatGPT Plus for one year, OpenAI announced on Saturday. The course, developed by the University of Malta, covers what AI is, what it can and cannot do and how to use it responsibly at home and in the workplace. “Malta is the first country to launch a partnership of this scale because we refuse to let our citizens stay behind in th…
Tokens on Solana that claim to track the private-market valuations of Anthropic and OpenAI dropped sharply this week after both AI companies warned that the structures used to back the tokens may be invalid.
OpenAI introduced a new cybersecurity initiative, Daybreak, on May 11, designed to find, validate, and help fix software vulnerabilities before attackers can exploit them. The firm describes the approach as making software “resilient by design,” moving security earlier into the build cycle through AI-assisted code review, threat modeling, patch validation, and dependency analysis. For crypto, …
A crypto founder had his laptop compromised when he joined what appeared to be a Microsoft Teams call with Pierre Kaklamanos, a Cardano Foundation contact he had spoken with before. When “Pierre” reached out about Atrium and sent a Teams invite, nothing looked out of place. On the call, the face and voice matched what …