OpenAI clamps down on security after foreign spying threats
Artificial intelligence group has added fingerprint scans and hired military experts to protect important data
OpenAI has overhauled its security operations to protect its intellectual property from corporate espionage, following claims of having been targeted by Chinese rivals.
The changes in recent months include stricter controls on sensitive information and enhanced vetting of staff, according to several people close to the $300bn artificial intelligence company.
The San Francisco-based start-up has been bolstering its security efforts since last year, but the clampdown was accelerated after Chinese AI start-up DeepSeek released a rival model in January.
OpenAI claimed that DeepSeek had improperly copied the California-based company’s models, using a technique known as “distillation”, to release a rival AI system. It has since added security measures to guard against these tactics.
DeepSeek has not commented on the claims.
The episode “prompted OpenAI to be much more rigorous”, said one person close to its security team, who added that the company, led by Sam Altman, had been “aggressively” expanding its security personnel and practices, including cyber security teams.
A global AI arms race has led to greater concerns about attempts to steal the technology, which could threaten economic and national security. US authorities warned tech start-ups last year that foreign adversaries, including China, had increased efforts to acquire their sensitive data.
OpenAI insiders said the start-up had been implementing stricter policies in its San Francisco offices since last summer to restrict staff access to crucial information about technologies such as its algorithms and new products.
The policies — known as information “tenting” — significantly reduced the number of people who could access the novel algorithms being developed, insiders said.
For example, when OpenAI was developing its new o1 model last year, codenamed “Strawberry” internally, staff working on the project were told to check that other employees were also part of the “Strawberry tent” before discussing it in communal office spaces.
The strict approach made work difficult for some staff. “It got very tight — you either had everything or nothing,” one person said. They added that over time “more people are being read in on the things they need to be, without being read in on others”.
The company now keeps a lot of its proprietary technology in isolated environments, meaning computer systems are kept offline and separate from other networks, according to people familiar with the practices. It also had biometric checks in its offices, where individuals could only access certain rooms by scanning their fingerprints, they added.
In order to protect model weights — parameters that influence how a model responds to prompts — OpenAI adopts a “deny-by-default egress policy”, meaning nothing is allowed to connect to the internet unless explicitly approved.
OpenAI had also increased physical security at its data centres, the people said. It was one of a number of Silicon Valley companies that stepped up their screening of staff and potential recruits because of an increased threat of Chinese espionage, the Financial Times reported last year.
Washington and Beijing are locked in a growing strategic competition, with the US imposing export controls to make it harder for China to obtain and develop cutting-edge technologies. However, concerns have also been raised about a rise in xenophobia at US tech companies given the prevalence of skilled workers of Asian descent.
OpenAI hired Dane Stuckey last October as its new chief information security officer from the same role at Palantir, the data intelligence group known for its extensive military and government work.
Stuckey works alongside Matt Knight, OpenAI’s vice-president of security products. Knight has been developing ways to use OpenAI’s large language models to improve its defences against cyber attacks, according to a person with knowledge of the matter.
Retired US army general Paul Nakasone was appointed to OpenAI’s board last year to help oversee its defences against cyber security threats.
OpenAI said it was investing heavily in its security privacy programs, as it wants to lead the industry. The changes were not made in response to any particular incident, it added.