AI firm Anthropic has announced a host of new security initiatives, including regular sweeps of the company’s premises, for both intruders and espionage devices.
Anthropic is one of the leading AI firms, quickly establishing itself both because of the strength of its models, as well as because of its focus on safe AI models. Its position in the industry makes it a target, however, and the company announced a number changes to deal with the growing threat it faces.
The company announced its plans in an update to it its Responsible Scaling Policy.
Physical security: Conduct regular Technical Surveillance Countermeasures (TSCM) at physical spaces using advanced detection equipment and techniques. Tailor TSCM sweeps to specific events, threats or incidents that may trigger an inspection. Regularly sweep physical premises for intruders and conduct physical security red-teaming.
Technical surveillance countermeasures are the industry term for efforts to find and neutralize surveillance devices. As the above highlights, the company is also planning to do regular sweeps of its premises to protect against intruders.
The Broader Picture
Anthropic’s updated security plans illustrate the importance of AI advances, and the competitive nature of the field, both in terms of competition between companies, as well as competition between countries.
The US was recently rattled by the emergence of DeepSeek, an AI model released by a relatively small Chinese company using third-rate hardware. Against all odds, DeepSeek easily competes with models like OpenAI’s ChatGPT and Anthropic’s Claude, despite being build with for a fraction of the cost.
What’s more, the US is working to cut off China’s access to advanced computer tech, especially tech that can be used in the development of advanced AI.
All of this combined is creating an environment where firms like Anthropic need to be as concerned about physical breaches as they are about cybersecurity ones.
from WebProNews https://ift.tt/iF2yXlC
No comments:
Post a Comment