After deep fakes, experts believe that Quantum AI can be the next threat that needs to be addressed. It is believed that Quantum computers have the potential to break currently used encryption methods, which rely on complex mathematical problems that are difficult for classical computers to solve. This could compromise sensitive information such as financial records, medical data, and government secrets. More so, data harvesting has become one of the pressing issues. This ‘harvest now, decrypt later’, strategy by Quantum AI can be a primary concern. “Quantum AI’s processing capabilities enable the development of more sophisticated and persuasive manipulations. This can potentially compromise certain encryption methods, leading to privacy breaches that indirectly affect manipulated content,” Sachin S. Panicker, chief AI scientist, Fulcrum Digital, told FE-TransformX.
Quantum AI – the threat of the present or future?
According to a 2022 survey, the adoption of quantum computing outpaced the adoption of artificial intelligence (AI) around the world, with about 49% of respondents reporting that they were adopting quantum more quickly than they did with AI. This trend was particularly in North America, with 62% of respondents highlighting a faster quantum adoption pace compared with AI, as per insights from Statista. According to a 2019 Forrester Research report, 80% of cybersecurity decision-makers expect AI to increase the scale and speed of attacks and 66% expected AI to conduct attacks that no human could conceive of.
Moreover, experts believe that quantum AI combined with a deep fake can play a lethal role in fraud. Recently, the CEO of a UK-based energy firm was conned out of $243,000 by scammers using deepfake AI voice technology to impersonate the head of the firm’s parent company. “Quantum AI takes deepfake threats to the next level, processing intricate algorithms at unprecedented speeds. This means more convincing and harder-to-detect deepfakes, posing a serious risk to consumers. Imagine audio deepfake scams that could exploit unsuspecting users,” Ritesh Chopra, India Director, Norton, highlighted.
With ‘harvest now, decrypt later,’ data and artificial intelligence’s ability to extract data, security breach is something that stands at risk. Consider extracting all minor data providing ‘signals intelligence’ and then combining it with Quantum AI, which could reveal intimate details about us individually, about our families, where we live and our health, among others and using them later on for fraudulent activities. “Quantum AI, a step ahead of deepfake technology, intensifies havoc potential. This dual threat has the potential to disrupt cross-border conflicts, especially in upcoming elections or conflicts like the Israel war, loom large, with Quantum AI manipulation influencing outcomes,” Ankur Srivastava, co-founder, CEO, QiMedia and QiTech, explained.
Coming to ‘identity’ risk, experts believe the combination of Quantum AI and deepfakes could create ‘fake’ individuals similar to real ones. Quantum AI is believed to have the potential to impact the digital world raising concerns regarding its role in geopolitical scenarios. In cross-border conflicts, Quantum AI-powered deepfakes could escalate tensions, leading to a more volatile international landscape.
Collaboration of threats: Quantum AI and AI
Industry experts believe that scammers with ‘harvest now, decrypt later’ collect and store encrypted data to decrypt it later, either by capitalising on advancements in today’s computers and new cryptographic attacks or utilising future quantum computers capable of breaking our current encryption standards. This, later on, might be used to create deepfakes for illegal activities such as creating misinformation and promoting fraudulent activities.
Quantum computing is believed to have shown the potential to generate more accurate predictive models on less training data. But quantum data can take on more than one state at a time, allowing quantum bytes to contain richer information. When applied to AI this allows to develop of much more complex models that are possible today with even the most advanced graphic processing unit hardware. This could result in hackers creating better-targeted content without needing to gather more data about their intended victims.
The road ahead
With the continuous advancements in technology, the question that remains unanswered is whether technology be used to fight against technology. It is believed that collaboration among digital engineering and business platform experts, policymakers, and industries establishes ethical frameworks, guiding the responsible development and deployment of emerging technologies, ensuring their positive societal impact while minimising potential harm. “ I believe that technology in collaboration with proper regulation can be an answer. Zero-trust-powered data loss prevention (DLP) can be used for securing misuse of artificial intelligence. User cases could include using ChatGPT for maintaining security while doing financial transactions,” Srikanth Nadhamuni, co-founder,Trustt, said.
Reportedly, Meta-owned Facebook AI Research (FAIR) and Inria( National Institute for Research in Digital Science and Technology), are developing a software called ‘stable signature’. Meta describes the labelling mechanism as an invisible watermarking technique created to distinguish when an image is created by an open-source generative AI model. “Deepfake (or indeed any other processing-intensive application) could benefit deeply from the processing jump that tech like quantum computing brings. However, the converse is also true – in an arms race where deepfakes are a big enough problem, enough resources (both monetary and computational, including the same quantum tech) would be deployed to counter the new threats generated,” Utkarsh Sinha, managing director, Bexley advisors, concluded.