What Using Security to Regulate AI Chips Could Look Like

Policy enforcement recommendations include limiting the performance of systems and implementing security features that can remotely disable rogue chips.
Governments have largely focused on software for AI policy, and the paper is a companion piece covering the hardware side of the debate, says Nathan Brookwood, principal analyst of Insight 64.
The industry will not welcome any security features that affect the performance of AI, he warns.
Throttling Connections Between Clusters One of the proposals the researchers suggest is a cap to limit the compute processing capacity available to AI models.
The idea is to put security measures in place that can identify abuse of AI systems, and cutting off and limiting the use of chips.
Specifically, they suggest a targeted approach of limiting the bandwidth between memory and chip clusters.
The easier alternative - to cut off access to chips - wasn't ideal as it would affect overall AI performance, the researchers wrote.
The paper did not suggest ways to implement such security guardrails or how abuse of AI systems could be detected.
Ways to limit network performance do exist for devices supporting the P4 programming language, which can analyze network traffic and reconfigure routers and switches.
Good luck asking chip makers to implement AI security mechanisms that could slow down chips and networks, Brookwood says.
Remote Possibilities Carry Some Risk The researchers also suggested disabling chips remotely, which is something that Intel has built into its newest server chips.
The On Demand feature is a subscription service that will allow Intel customers to turn on-chip features such as AI extensions on and off like heated seats in a Tesla.
The researchers also suggested an attestation scheme where chips allow only authorized parties to access AI systems via cryptographically signed digital certificates.
Firmware could provide guidelines on authorized users and applications, which could be changed with updates.
While the researchers did not provide technical recommendations on how this would be done, the idea is similar to how confidential computing secures applications on chips by attesting authorized users.
Intel and AMD have confidential computing on their chips, but it is still early days yet for the emerging technology.
There are also risks to remotely enforcing policies.


This Cyber News was published on www.darkreading.com. Publication date: Fri, 16 Feb 2024 22:25:10 +0000


Cyber News related to What Using Security to Regulate AI Chips Could Look Like

What Using Security to Regulate AI Chips Could Look Like - Policy enforcement recommendations include limiting the performance of systems and implementing security features that can remotely disable rogue chips. Governments have largely focused on software for AI policy, and the paper is a companion piece ...
11 months ago Darkreading.com
TPM Chips and the Use of TPM in Virtualization Technology - TPM chips have grown in relevance in both physical and virtual contexts, where they play a critical role in data security and preserving the integrity of computer systems. TPM chips, their functionality, and how they are used in virtualization ...
1 year ago Feeds.dzone.com
Here's Why the World is Investing So Much in Semiconductors - Hannah Mullane, a BBC correspondent, recently visited Pragmatic Semiconductor, the UK's newest computer chip facility in Durham. The large site is being turned into a sophisticated computer chip production hub. Pragmatic Semiconductor has already ...
1 year ago Cysecurity.news
China Telecom Trains AI Model Using Domestic Chips - The Institute of AI at China Telecom, one of China’s main state-backed telecoms firms, said the open source TeleChat2-115B and a second unnamed model were trained using tens of thousands of domestically produced chips. ByteDance, the ...
4 months ago Silicon.co.uk
9 Best DDoS Protection Service Providers for 2024 - eSecurity Planet content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More. One of the most powerful defenses an organization can employ against distributed ...
1 year ago Esecurityplanet.com
US Commerce Review Semiconductor Supply Chain - National security move, as the US Commerce Dept says it will launch survey of US semiconductor supply chain. The US Department of Commerce has announced a new review, in a sign of the continuing geopolitical tensions around the world, and the ...
1 year ago Silicon.co.uk
Apple's AI Moves Will Impact Future Chip, Cloud Security Plans - The measures Apple has implemented to prevent customer data theft and misuse by artificial intelligence will have a marked impact on hardware security, especially as AI becomes more prevalent on customer devices, analysts say. Apple emphasized ...
7 months ago Darkreading.com
US To Invest $5 Billion In Semiconductor Research - White House confirms $5 billion from US Chips and Science Act will be used for semiconductor research and development. The Biden Administration has confirmed that the United States is setting aside billions of dollars for semiconductor-related ...
11 months ago Silicon.co.uk
Nvidia To Build Network Of AI Chip Plants In Japan - Nvidia chief Jensen Huang says company to work with local companies to build network of AI chip plants in Japan. Nvidia is to collaborate with local companies to build a network of semiconductor manufacturing facilities in Japan to meet demand for ...
1 year ago Silicon.co.uk
Microsoft Security Copilot improves speed and efficiency for security and IT teams - First announced in March 2023, Microsoft Security Copilot-Microsoft's first generative AI security product-has sparked major interest. With the rapid innovations of Security Copilot, we have taken this solution beyond security operations use cases ...
1 year ago Microsoft.com
Cybersecurity jobs available right now: October 2, 2024 - Help Net Security - As an Applied Cybersecurity Engineer (Center for Securing the Homeland), you will apply interdisciplinary competencies in secure systems architecture and design, security operations, threat actor behavior, risk assessment, and network security to ...
4 months ago Helpnetsecurity.com
2024 brings changes in data security strategies - 2024 will be a revolutionary year for the data security landscape as Data Security Posture Management technology rapidly evolves to keep pace with the colossal amount of data being created, stored and shared within organizations and across business ...
1 year ago Helpnetsecurity.com
What Is Cloud Security Management? Types & Strategies - Cloud security management is the process of safeguarding cloud data and operations from attacks and vulnerabilities through a set of cloud strategies, tools, and practices. The cloud security manager and the IT team are generally responsible for ...
8 months ago Esecurityplanet.com
Embracing Security as Code - Everything is smooth until it isn't because we traditionally tend to handle the security stuff at the end of the development lifecycle, which adds cost and time to fix those discovered security issues and causes delays. Over the years, software ...
1 year ago Feeds.dzone.com
10 Best Security Service Edge Solutions - Security Service Edge is an idea in cybersecurity that shows how network security has changed over time. With a focus on customized solutions, Security Service Edge Solutions leverages its expertise in multiple programming languages, frameworks, and ...
11 months ago Cybersecuritynews.com
What is Security Posture? - Security posture is a term often mentioned in cybersecurity, with businesses often told to improve or maintain a robust security posture. With the onset of 2024, now is a better time than ever to take stock of your company's security posture and plan ...
1 year ago Securityboulevard.com
US Moves To Facilitate AI Chip Shipments To Middle East | Silicon - The rule will allow data centres in the region to apply for Validated End User status, under which US tech companies will be authorised to sell advanced chips to the firms under a general authorisation. Shipments of advanced artificial intelligence ...
4 months ago Silicon.co.uk
Five business use cases for evaluating Azure Virtual WAN security solutions - To help organizations who are evaluating security solutions to protect their Virtual WAN deployments, this article considers five business use cases and explains how Check Point enhances and complements Azure security with its best-of-breed, ...
8 months ago Blog.checkpoint.com
Strengthening Security Posture Through People-First Engagement - Regular, small doses of security education help combat the “forgetting curve,” a theory developed by Hermann Ebbinghaus that suggests people forget 75% of newly learned information within a couple of days. These statistics underscore a critical ...
4 months ago Informationsecuritybuzz.com
Managing the Risk of Cancer in Security - Recently, a friend brought up the term Carcinization and I had to look it up. It turns out that this term was created more than a century ago to explain the process of crustaceans transforming into crab-like forms. What does this example of ...
2 years ago Securityweek.com
6 Best Cloud Security Companies & Vendors in 2024 - Cloud security companies specialize in protecting cloud-based assets, data, and applications against cyberattacks. To help you choose, we've analyzed a range of cybersecurity companies offering cloud security products and threat protection services. ...
11 months ago Esecurityplanet.com
Intel knew AVX chips were insecure and did nothing - Intel has been sued by a handful of PC buyers who claim the x86 goliath failed to act when informed five years ago about faulty chip instructions that allowed the recent Downfall vulnerability, and during that period sold billions of insecure chips. ...
1 year ago Theregister.com
Normalizing Security Culture: Stay Ready - While it may seem like self-promotion or extraneous work, it’s extremely valuable to take the extra time to summarize threats stopped, processes improved, projects completed and team members modeling strong security behavior. Most people don't ...
4 months ago Darkreading.com
Mastering SDLC Security: Best Practices, DevSecOps, and Threat Modeling - In the ever-evolving landscape of software development, it's become absolutely paramount to ensure robust security measures throughout the Software Development Lifecycle. Each of these have illuminated different vulnerabilities that can be exploited ...
1 year ago Securityboulevard.com
Benefits and challenges of managed cloud security services - Too many organizations lack the in-house cloud security expertise and resources needed to protect cloud assets effectively. One option to address these challenges is managed cloud security. Outsourcing cloud security to a third party not only helps ...
11 months ago Techtarget.com

Latest Cyber News


Cyber Trends (last 7 days)


Trending Cyber News (last 7 days)