-
China's memories of Mao fade to nostalgia 50 years after his death
-
Messi creates both Miami goals in 2-2 draw with Atlanta
-
US envoys discuss with Putin 'substantive' plans for steps to end Ukraine war
-
Rybakina books US Open fourth-round berth, inches closer to No. 1
-
Zheng stuns Keys as US Open upsets dim American hopes
-
Sobering reality: Curbs on alcohol in northeast Syria cause discontent
-
Jovic 'loses sanity five times over' to oust Eala and book Gauff clash
-
Wobbly Osaka 'shocked' to battle past Mertens and reach last 16
-
'Like a jewel': Japan's luxury ice revered by bars worldwide
-
Indian Oscar-nominated filmmaker turns camera on his cancer
-
'The flood is coming': How a Nepal principal saved 900 students
-
Argentina 'let off hook' by Australia as Kiss's winning run ends
-
Kiss's winning run ends as Argentina hold Australia to draw
-
Argentina threatens to sanction oil projects around Falklands
-
Alcaraz excited to face 'dangerous' Paul for US Open quarter-final berth
-
Gauff into round four at US Open after straight-set win
-
Prince William to represent King Charles III at Norway late King Harald's funeral
-
Bordeaux hammer Racing in Top 14 start as Bielle-Biarrey extends deal
-
Inter and Roma roar back to win thrillers and share top spot
-
Zheng, Potapova pull off upsets to reach US Open last 16
-
Don't stop them now: fans party for Freddie Mercury's 80th birthday
-
Swiatek into US Open fourth round in tight two-setter
-
AFP photographer wins top prize for Madagascar protest coverage
-
Werro wins in Brussels, as Russell and Dos Santos shine
-
Keys: 'I felt like I couldn't play tennis'
-
Kolbe breaks jaw and shoulder injury sidelines Savea
-
Simmonds stars as Pau upset Montpellier
-
US military strikes oil tankers as Iran vows tougher reprisals
-
Inter roar back in five-goal thriller with Napoli
-
Zheng rallies to stun Keys, Anisimova out in US Open third round
-
Haaland keeps Man City perfect in Premier League, blunt Spurs held
-
Bayern held to draw by stubborn Schalke
-
Hamilton worried by Ferrari pace while Antonelli prepares for a fight
-
Lens lose again ahead of Champions League date
-
Vuelta stage 15 shortened due to extreme heat, Mas consolidates lead
-
Olympic champion Zheng storms home to reach US Open last 16
-
Putin meets US envoys for Ukraine war talks
-
Big-spending Spurs must make better attacking decisions, says De Zerbi
-
Modric available for Croatia in Nations League
-
South Africa dominate New Zealand in second half to lead series
-
Potapova upsets 2025 US Open runner-up Anisimova
-
Kiefer Sutherland still in action, but ibuprofen lending helping hand
-
Gasly a 'class act' says Alpine boss after Monza pole
-
'Are we ready?' Oasis film captures band's frosty reunion
-
Athletic Bilbao crush Simeone's flat Atletico
-
Commemoration held for late Bosnian Serb war criminal Ratko Mladic
-
Stay far from battlefield, warns Japanese director in Vietnam War film
-
Haaland keeps new-look Man City perfect in Premier League, blunt Spurs held
-
Tens of thousands of Croatians protest over hazardous waste
-
Alpine's Gasly takes shock pole for Italian Grand Prix
OpenAI to launch new model with 'stronger safeguards' after hack
ChatGPT maker OpenAI said Tuesday it was preparing to release its newest powerful model, known as Astra, after implementing "stronger safeguards" following a rogue cyberattack involving a different AI model.
The San Francisco-based artificial intelligence (AI) giant paused some of its model development for two weeks this summer after two models it was testing were involved in a security breach of software company Hugging Face.
Although Astra "was not involved" in the incident, OpenAI has beefed up its safety measures, the company said in a blog post.
"We have since implemented even stronger safeguards for Astra, including training the model to more reliably refuse harmful cyber requests and respect safety restrictions, additional protections against misuse, and monitoring that can stop potentially unauthorized activity," the blog said.
That includes classifying Astra as reaching a "critical cybersecurity threshold," which means OpenAI believes the model is capable of finding and exploiting cybersecurity gaps.
"It is the first model we are designating at this level, and requires stronger safeguards during development and before release," the blog said.
When OpenAI eventually launches Astra, access to certain capabilities will be limited and the most advanced capabilities will be made available to a select group of early testers, the blog said.
Concerns have increased in recent months about the capabilities of advanced AI models after incidents involving models from both OpenAI and rival developer Anthropic, though none of the models in those incidents were available to customers.
Anthropic also recently discovered that its models had gained unauthorized access to three unnamed organizations during testing that was supposed to keep them away from "real-world" systems.
Last week, more than 100 organizations around the world, including OpenAI and Anthropic, signed an open letter calling for a global effort to "strengthen cyber defenses" against AI-powered cybersecurity threats.
"We have a limited window to strengthen cyber defenses," the letter said. "In the coming months, AI-enabled cyber attacks will become far more widespread and sophisticated as models around the world become increasingly capable."
O.Norris--AMWN