UC Round Table: The Future of Voice

"We're making strides in the battle against robocalls and scams, showing promising results," Lenman explained. "Now, our focus shifts to restoring faith in the voice channel. It's essential for end-users to trust the authenticity of their callers, ensuring engagements are secure and personal information is safeguarded. Achieving this means having stringent controls over call originators, verifying brands, and guarding against spoofing."

"By reinforcing the security foundations of voice services — a channel we've relied on for decades — we're committed to rebuilding trust," Lenman continued. "This effort is crucial for protecting voice data with robust encryption and compliance with privacy regulations, ensuring a secure and transparent communication environment. "

Joel Keene, Product Manager at Continuant

For Keene, the era of AI and machine learning complicates security.

"It means a couple of things," Keene outlined, "the number of security attacks has vastly increased over anything we’ve seen in the past, and the exploitation of readily available Internet data sources can be correlated quickly by LMs that can be designed to determine identity."

Traditionally, organisations think in terms of protecting data connections, but the voice networking aspect is significant, Keen suggested.

"AI models can make a caller’s voice indistinguishable from a mimicked source," Keene said. "Think in terms of an analyst receiving a call from an AI that sounds exactly like their manager or even the CEO giving them instructions or guidance. The ability to determine calling source and identity will be critical in minimising business interruptions and hacking attempts through social engineering."

"These services can be cost-prohibitive and are almost all administratively complex to maintain outside of a CPaaS solution like Continuant Connect."

David Keedy, Microsoft Sales Architect at Continuant

Keedy stated that, specifically, that voice security (toll fraud, DDOS) is always important.

"The ability to integrate AI into existing platforms protecting customers' services will be invaluable to increase detection and mitigate risks more quickly," Keedy elaborated. "Also, STIR/SHAKEN protocols will be improved upon by AI as it can police and update these systems more effectively in real time."

George Schoenstein, CMO at Fusion Connect

Schoenstein raised several significant security and privacy concerns.

"Interception threats and unauthorised access to voice conversations pose risks, necessitating robust encryption protocols," Schoenstein said. "Organisations must prioritise end-to-end encryption to safeguard sensitive information.

"The utilisation of voice recognition technology introduces additional privacy considerations, particularly regarding biometric data. Voiceprints, a form of biometric data used for authentication, are inherently unique to individuals, amplifying privacy concerns if mishandled. Unauthorised access to such biometric data can lead to severe consequences, including identity theft."

Schoenstein affirmed that, therefore, it's imperative for organisations to establish stringent privacy policies and ensure compliance with relevant data protection regulations to uphold individuals' privacy rights effectively.

"To address these challenges, organisations should implement multi-factor authentication to enhance access controls," Schoenstein stated. "Regular security audits and updates are crucial for identifying and mitigating vulnerabilities. Transparent communication about data usage policies with users fosters trust and compliance. Furthermore, investing in secure cloud infrastructure and adopting decentralised storage options can enhance data protection."

"Collaborating with cybersecurity experts and staying abreast of evolving threats ensures that organisations proactively tackle security and privacy concerns associated with the burgeoning use of voice data in UC systems."

What are the potential benefits and challenges of using AI-driven voice features in UC platforms?

Joel Keene, Product Manager at Continuant

Keene tied his answer to the latter aspect of the question to his previous response about security and privacy — and the potentially seismic impact AI will play in it.

However, for the benefits, Keene suggested being able to "manage voice solutions and integrations with simple language (something like asking Alexa to move phone numbers from one user to another)".

"On a larger scale, front-ending calls in either direction with a Conversational AI offer a lot of potential in minimising the need for extensive contact centres that may be focused on script-driven conversation techniques," he added.

David Keedy, Microsoft Sales Architect at Continuant

Keedy's purported benefits include a platform with more capabilities and the ability to implement without needing specialised education and training.

"The challenge," Keedy continued, "will be in shaping the AI product to work with a wide variety of customers and be able to discern what the customer wants/needs to do. It would be great if AI could be predictive and suggest scenarios, allowing the end user to pick what works best for them."

Anders Lenman, Vice President, RTCx Product Operations & Management at Sinch

Lenman argued that AI-driven voice features are revolutionising UC platforms, enhancing efficiency, customer service, and data analysis accuracy.

"However," he contextualised, "ensuring the reliability and fairness of AI interpretations, as well as maintaining privacy and security, are challenges that need attention. As AI becomes a bigger part of UC platforms, we're focused on ethical AI use and constantly improving AI algorithms. "

George Schoenstein, CMO at Fusion Connect

Schoenstein pinpointed that AI enhances user experiences by allowing natural language processing, facilitating more intuitive interactions. AI-driven voice features can automate routine tasks, boosting efficiency and productivity.

"Predictive analytics, enhanced by AI, significantly bolsters the capacity for informed decision-making within UC systems," Schoenstein continued. "By harnessing vast data streams and employing advanced algorithms, AI can forecast future trends, behaviours, and potential issues with remarkable accuracy. This empowers UC systems to preemptively address challenges, optimise resource allocation, and tailor services to meet evolving user needs, converting your voice data into a critical business asset."

However, Schoenstein added that AI’s ability to process voice data for analysis can pose a safety risk, meaning striking a balance between personalisation and privacy is essential. "Another challenge is the potential biases embedded in AI algorithms, which can affect voice recognition accuracy and may lead to unintended consequences," he said. "Ensuring robust cybersecurity measures is pivotal to protecting against AI-related vulnerabilities."