29 décembre 2024 | International, Aérospatial
14 février 2019 | International, C4ISR
What the Pentagon’s new AI strategy means for cybersecurity
By: Justin Lynch
The new Pentagon artificial intelligence strategy serves as a roadmap for how the American military will embrace machine learning in future cyber operations, which could be a boon for companies who invest in the technology.
Released Feb 12, the Pentagon's strategy shows how the American military will rely on artificial intelligence as a defensive tool.
“We will increase our focus on defensive cybersecurity of hardware and software platforms as a precondition for secure uses of AI,” the Pentagon's new strategy states. “ In order to ensure DoD AI systems are safe, secure, and robust, we will fund research into AI systems that have a lower risk of accidents; are more resilient, including to hacking and adversarial spoofing.”
A particular focus of the new strategy is researching “emergent effects,” which is what happens when two artificial intelligence systems interact. The Army's Aberdeen Proving Ground, for one, is conducting studies into how artificial intelligence can protect soldiers' tactical networks and communications from cyberattacks.
The Pentagon's artificial intelligence strategy comes as both the American military and hackers have embraced machine learning for cyberattacks.
According to an analysis by Fifth Domain, more than half of the new challenges and programs announced by the intelligence community's research arm in 2018 involve machine learning or predictive analytics, which represents a potential boon for cybersecurity researchers.
In December, the Pentagon announced that Air Force Lt. Gen. Jack Shanahan would be head of a new artificial intelligence center.
Hackers are using artificial intelligence during cyberattacks, Sam Curry, the chief security officer at Cybereason, a threat intelligence firm, told Fifth Domain. Curry said that a particular risk with artificial intelligence is the use of “second order chaotic systems,” where events cause data inputs to change, using the stock market as an example.
“Second order systems with badly applied, overreacting machine learning are prone to exploitation,” Curry said.
With the release of the artificial intelligence strategy, the Pentagon is attempting to combat significant investments in machine learning by other countries.
“Other nations, particularly China and Russia, are making significant investments in AI for military purposes, including in applications that raise questions regarding international norms and human rights,” the strategy says. “The costs of not implementing this strategy are clear. Failure to adopt AI will result in legacy systems irrelevant to the defense of our people, eroding cohesion among allies and partners, reduced access to markets that will contribute to a decline in our prosperity and standard of living, and growing challenges to societies that have been built upon individual freedoms.”
Sur le même sujet
-
-
9 novembre 2020 | International, Aérospatial
More funding endorsed for Evader aerial target missile system
by Julian Kerr Additional funding has been endorsed, although not yet approved, to confirm the supersonic performance of the Australian-developed Evader autonomous aerial target missile system, sources close to the programme have disclosed. An application for a grant of AUD2.85 million (USD2.02 million) from the government's Defence Innovation Hub has been endorsed by both the Royal Australian Navy (RAN) and the Australian Army to fund a second series of flight trials for the low-cost, air-launched system, said the sources. A decision is expected shortly. Initial flight trials, funded by the Innovation Hub and carried out in 2018 at South Australia's Woomera range, validated the Evader's design, autonomous control system and stainless steel ramjet, and proved the parachute-recoverable, 4.2 m-long target missile could be re-used up to 10 times. Endurance of the 90 kg missile is about 30 minutes. While the initial trials were subsonic, the second series of trials will be low-supersonic following relocation of the variable air intake from the top to the bottom of the missile's fuselage. Speeds of up to Mach 2 are anticipated in subsequent trials, with future variants intended to reach Mach 3. https://www.janes.com/defence-news/news-detail/more-funding-endorsed-for-evader-aerial-target-missile-system
-
5 octobre 2020 | International, Aérospatial
Pentagon's Lord blames KC-46A's fixed-price contract structure for further programme issues
Pat Host The Pentagon's top acquisition executive blames the Boeing KC-46A Pegasus aerial refuelling tanker's fixed-price contract structure for the programme's ongoing issues. Senator Jeanne Shaheen of New Hampshire said during a 1 October hearing that Boeing cancelled a 2 October KC-46A delivery to the 157th Air Refueling Wing (ARW) in New Hampshire due to electrical problems with the aircraft. Ellen Lord, under secretary of defense for acquisition and sustainment (A&S), said in response that the root cause of the KC-46A's problems is the contract type that was awarded to Boeing. Lord said that one issue is the technical solution that was the aircraft's original design, which is now being redesigned. The programme, she said, also has a myriad of manufacturing issues including ongoing foreign object debris (FOD) problems. The programme, Lord said, has an engineering design and execution issue as well as a manufacturing problem. Boeing's fixed-price contract structure, previously lauded as an example of improved government contracting, means the company pays the Pentagon for cost overruns. The contract structure was considered an improvement over cost-plus contracting, where the government reimburses the contractor for cost overruns. One expert said that Boeing paying for cost overruns on the KC-46A disincentivises the company from improving its performance. Doug Birkey, Mitchell Institute for Aerospace Studies executive director, told Janes https://www.janes.com/defence-news/news-detail/pentagons-lord-blames-kc-46as-fixed-price-contract-structure-for-further-programme-issues