Artificial Intelligence and the Future of Hearing Aids

Artificial intelligence (AI) is now transforming hearing aids, turning them from simple sound amplifiers into intelligent devices that adapt to real-life situations. At its core, AI refers to computer systems that can learn from data and make decisions or predictions, rather than just following fixed instructions.

In healthcare, AI is already being applied in areas like medical imaging and drug discovery. In hearing technology, it is opening the door to significant advances: clearer speech in noisy environments through deep neural network noise reduction, personalised sound adjustments based on user preferences, proactive monitoring of device performance, and app-based personal assistants that let wearers fine-tune their hearing in real time. Together, these innovations are helping people not only hear better but also enjoy a more personalised, reliable, and connected listening experience.

Why Background Noise Is Such a Problem

Globally, more than 430 million people live with disabling hearing loss. While hearing aids provide significant benefits, they sometimes fall short in environments with background noise—such as restaurants, social gatherings, or busy workplaces.

Most hearing aids rely on directional microphones and noise-reduction algorithms to improve the signal-to-noise ratio (SNR). These can work well when the source of interest is detected immediately in front of, to the side of, or behind the listener. However, they struggle when:

• Multiple talkers are present.
• Important sounds originate from outside the microphone beam (e.g., from various directions simultaneously).
• Background noise is unpredictable.

As The Hearing Review explains, even advanced beamforming systems “require the user to have the awareness and ability to place the signal of interest within the beam,” which can be impractical in real-world settings. The result is listening fatigue, reduced communication, and, in many cases, reduced hearing aid use.

How AI Helps: Smarter Noise Reduction

One of the biggest challenges for people using hearing aids is separating speech from background noise. Traditional systems, such as Active Noise Cancellation (ANC), mainly block steady, low-frequency sounds like the hum of an air conditioner. However, they may struggle in busy, unpredictable environments such as restaurants or shopping centres.

This is where artificial intelligence (AI) and deep learning play a role. Deep learning is a type of AI that mimics how the brain learns from experience. Instead of following fixed rules about what is “speech” and what is “noise,” deep neural networks (DNN) are trained on massive collections of real-world sounds. This helps them recognise and separate voices from unwanted noise with much greater accuracy.

Hearing aid technology has made advances in this area:

Selective Noise Cancellation (SNC): Unlike traditional ANC, SNC can distinguish speech from background chatter—even in echoey or noisy environments. SNC works by analysing the unique patterns of speech and treating them differently from other sounds, so a friend’s voice can be highlighted. At the same time, clinking dishes or overlapping conversations are pushed into the background.

Real-time processing: New designs, such as Convolutional Recurrent Networks (CRNs) and Transformers, can perform this separation in under 10 milliseconds, which is fast enough to maintain a natural conversation. CRNs combine the strengths of two types of AI—convolutional networks for recognising sound features and recurrent networks for tracking how those sounds change over time—making them especially good for fast-changing noise environments like busy streets or cafés.

Improved clarity: Clinical studies have shown that AI-based systems can enhance speech clarity by up to 18.3 dB SI-SDR compared to older methods. SI-SDR (Scale-Invariant Signal-to-Distortion Ratio) is a technical measure of how clearly speech can be separated from noise; a boost of 18.3 dB represents a dramatic improvement, similar to turning down the volume of background noise while turning up the clarity of a person’s voice at the same time.

By making speech clearer and reducing background distractions, these AI systems also lower the mental effort needed to listen. Over time, this can help reduce listening fatigue and support better long-term brain health.

Imagine sitting at a family dinner. With older technology, the laughter and overlapping conversations might blur together, making it hard to keep up. With AI-driven selective noise cancellation and real-time processing, your hearing aids can bring your cousin’s voice into focus—even while the kids are chattering and dishes are clattering in the background.

Personalisation Through AI-Enabled Apps

The benefits of AI go beyond the hearing aid itself. Many modern devices now connect to smartphone apps that act as “AI assistants.” These apps serve two essential roles:

User feedback loop – Instead of visiting a clinic every time adjustments are needed, wearers can provide real-time feedback via the app. This data helps the AI learn the user’s personal preferences—whether they prefer more background reduction in cafés or softer amplification at home.

Context awareness – Apps can combine hearing aid data with smartphone sensors (such as location and noise environment) to improve context-based adjustments. Over time, this creates a more personalised listening experience tailored to the individual.

DNN Noise Reduction and Hardware Challenges

As The Hearing Review points out, deep neural network (DNN) noise reduction is already widely used in everyday consumer applications, such as video conferencing and online meetings. In this context, a DNN is trained on massive amounts of audio data to recognise patterns in speech and noise. When you’re on a call, the system can filter out the hum of an air conditioner or the clatter of a keyboard while keeping your voice crisp and clear.

Applying this same technology to hearing aids has enormous potential. Instead of treating all sound equally, DNN-based noise reduction can learn to separate voices from background noise in real-time, even in complex environments such as restaurants or busy streets. This means wearers don’t just get “louder” sound—they get cleaner, more natural conversations with less mental strain.

However, implementing DNN technology in hearing aids comes with unique challenges. Consumer applications, such as video conferencing software, run on laptops, smartphones, or cloud servers that have virtually unlimited processing power and battery capacity. Hearing aids, by contrast, are tiny devices that must fit discreetly behind or inside the ear. 

Battery Limits: Hearing aids use tiny batteries—often rechargeable cells smaller than a fingernail. Running a high-powered DNN continuously can quickly drain these resources, resulting in shorter listening times for users.

Processing Power: Complex AI models can involve millions—or even billions—of parameters, requiring chips capable of lightning-fast calculations. Shrinking that level of computing into a chip small enough to fit inside a hearing aid is a significant engineering challenge.

Heat and Size Constraints: Unlike phones or laptops, hearing aids have no space for cooling systems. Chips must therefore be both energy-efficient and heat-efficient, while still delivering real-time results.

In other words, it’s like trying to fit a supercomputer into something the size of a jellybean—all the power has to be there, but in miniature, without draining the battery or overheating.

Recent breakthroughs in chip design are starting to bridge this gap. Manufacturers are now developing specialised low-power AI processors—such as Phonak’s DEEPSONIC and other dedicated neural processing units—that can run advanced DNN algorithms while staying small and energy-efficient enough for daily wear. 

The Future of Hearing

Companies like Widex, Phonak, Starkey, and Signia are already integrating AI-powered selective noise cancellation into their devices. As this technology advances, the focus will increasingly shift from amplifying sound to delivering comfortable, personalised communication experiences.

For New Zealanders living with hearing loss, these advances mean:

• Better performance in noisy environments.
• Less listening effort and fatigue.
• More control and personalisation through easy-to-use apps.
• Improved reliability and fewer disruptions.

AI isn’t replacing hearing care professionals—it’s enhancing their role. By combining innovative technology with clinical expertise, hearing aid wearers can enjoy a more natural, connected listening experience than ever before.

Signia: AI Assistant in Your Pocket 

Signia has taken AI straight to the user with the Signia Assistant, built into their smartphone app. Instead of waiting for your next clinic appointment to fine-tune your hearing aids, you can ask the Assistant for help in the moment.

How it works: If you’re having trouble in a noisy environment, you open the app and activate the Signia Assistant. The AI suggests a new setting to improve speech clarity. You can accept, reject, or tweak the change, and the system learns from your choice.

Learning over time: The Assistant combines your feedback with anonymised data from thousands of other users worldwide. This helps the system become more adept at solving common listening problems.

Teamwork with professionals: All of your adjustments are logged and shared with your hearing care professional through Signia’s Connexx fitting software. That way, your clinician can see what changes you’ve tried and refine your settings further if needed.

The result is a highly personalised, real-time tuning system that puts the user in control while keeping professionals informed.

Phonak: AI-Powered Clarity and Reliability

Phonak’s latest AI innovation is the DEEPSONIC chip, designed specifically for hearing aids. This dedicated processor powers Spheric Speech Clarity, which can separate speech from noise in every direction—not just in front of the wearer. In tests conducted by Phonak, this technology improved speech understanding by more than 36% compared to other leading devices, providing wearers with a clearer, more natural listening experience in challenging environments.

Phonak is also focusing its AI efforts on device reliability, making hearing aids as seamless and worry-free as possible:

Proactive monitoring: Phonak hearing aids collect performance data—such as reboot frequency, battery behaviour, and system irregularities—in real time.

AI insights: This data is analysed using AI to identify patterns and predict potential issues before they affect the wearer. For example, devices that reboot frequently can be flagged for early service.

Fewer disruptions: By addressing problems through firmware updates or replacements before the user notices them, Phonak helps reduce unnecessary troubleshooting appointments and service calls.

With DEEPSONIC delivering powerful sound clarity and AI-driven monitoring ensuring reliability, Phonak is combining hearing performance with peace of mind for both wearers and hearing care professionals.

Widex: Real-Time Machine Learning with SoundSense Learn

Widex has taken a slightly different path, focusing on machine learning—a branch of AI that enables systems to continually improve based on feedback. Their SoundSense Learn feature, available in the Widex EVOKE hearing aids, lets users personalise sound directly from their phone.

Interactive adjustments: Through a simple app interface, wearers are asked to choose between different sound options—“Does A or B sound better?” In just 20 quick steps (as opposed to millions of possible combinations), the system determines the optimal setting for that person in that situation.

Workplace advantage: Data shows that many users create custom programs for the workplace, highlighting the vital importance of real-time adjustments for effective communication on the job.

User empowerment: Instead of trying to remember how a difficult listening situation felt and explaining it at the next appointment, users can fix it on the spot.

Widex’s approach combines the expertise of hearing professionals with the power of real-time learning, ensuring every user can fine-tune their devices to match their unique needs.

Starkey: A Holistic Approach with Health and Safety Features

Starkey takes AI in hearing aids beyond just sound, blending hearing support with overall health and well-being. Their flagship devices, including the Genesis AI and Edge AI, are designed to mimic the way the brain naturally processes sound while also monitoring physical health and safety.

Neuro Sound Technology: Starkey’s AI-powered processor classifies complex soundscapes in real time, enhancing speech clarity while reducing distracting background noise.

Fall detection: Built-in 3D motion sensors can detect falls and automatically alert pre-selected family members or caregivers, offering added safety for older users.

Activity and wellness tracking: Steps, engagement levels, and activity goals can all be monitored through the My Starkey app, enabling users to track both their hearing health and overall well-being.

Starkey’s approach is holistic—it’s not just about hearing better, but about supporting independence, safety, and overall health. For many users, this makes the device feel like both a hearing aid and a personal health companion.

Final Thoughts

Artificial intelligence is no longer just a buzzword—it’s becoming a practical tool that is reshaping hearing care. By moving beyond simple amplification, AI-powered hearing aids are learning to understand real-world environments, separate speech from noise, and adapt to individual preferences in ways that traditional technology cannot.

Each manufacturer has taken its own approach:

• Signia gives users direct control with the Signia Assistant app for real-time, personalised adjustments.
• Phonak focuses on reliability and advanced sound clarity through proactive monitoring and its DEEPSONIC AI chip.
• Widex empowers wearers with real-time machine learning, enabling instant fine-tuning in any situation.
• Starkey blends hearing support with health and safety features, offering fall detection, wellness tracking, and holistic care.

For individuals with hearing loss, these innovations will hopefully enable clearer conversations, reduce listening fatigue, and increase independence. Importantly, AI doesn’t replace hearing care professionals—it works alongside them. The result is a new era of hearing solutions that are smarter, more reliable, and better suited to the challenges of everyday life.

As AI continues to evolve, hearing aids will become even more intuitive, offering not just better hearing but also improved well-being and peace of mind. The future of hearing care is intelligent, personalised, and designed to help people stay connected—wherever life takes them.