Read time: 7 minutes
Highlights
Sweetness is becoming a trust problem, not just a sugar problem.
- A new concern: Fresh research on sorbitol is opening questions about how some sugar-reduction ingredients behave in the body.
- A different architecture: Sweet proteins could deliver intense sweetness in tiny amounts, but they won't replace sugar's bulk, structure or browning on their own.
- A growing ecosystem: Oobli, Amai Proteins, Nambawan Spain, Sweegen, Bestzyme and Pentasweet are signalling that the category is moving towards commercial scale.
- A strategic test: The winners won't simply remove sugar. They'll redesign the whole food matrix and communicate with far greater transparency.
Is the confectionery industry ready for the next sugar-reduction conversation?
I don't think it has a choice!
For years, we treated sugar reduction as a substitution exercise. Take out the sugar. Add a sweetener. Adjust the flavour. Move on. Sweet proteins are now challenging that logic.
But new reporting from ConfectioneryNews shows why that approach is becoming harder to defend. Consumers aren't only asking how much sugar is in a product anymore. They're asking what replaced it, how it works and whether the new ingredient has a credible long-term story.
This is where sweet proteins become fascinating. They could become part of a new confectionery architecture built around lower sweetness, better food matrices and highly potent ingredients used in tiny amounts.
The trust problem behind sugar reduction
The Washington University in St Louis research reported by ConfectioneryNews focuses on sorbitol. Sorbitol is a type of sweetener made from sugar, known as a sugar alcohol. It isn't the same as ordinary table sugar. It's used in products including confectionery, baked goods and sugar-reduced foods because it can provide sweetness, bulk and moisture control.
The research raises a concern about how sorbitol may be processed under particular conditions. In animal experiments, researchers found that sorbitol could be turned into compounds that behave much like fructose, the natural sugar found in fruit, especially when gut bacteria couldn't break it down effectively.
The research doesn't prove that ordinary human consumption of sorbitol causes liver disease. The study used fish and concentrated exposure. The researchers have also said that more human research is needed. The issue is better understood as a research question, not a finished health conclusion.
Fatty liver disease, which doctors now call metabolic dysfunction-associated steatotic liver disease, or MASLD for short, is a condition where too much fat builds up in the liver.
It has many possible drivers. We shouldn't turn one animal study into a sweeping claim about every sugar-reduction ingredient.
At the same time, brands can't ignore the direction of travel.
The same article discusses separate associations involving xylitol, neotame, sucralose and multiple sweeteners. Again, these are links, not proof that one causes the other. But every new headline adds pressure to a category that was once presented as an uncomplicated solution.
If the industry keeps presenting reformulation as a numbers game, consumers may reasonably ask whether the new formula is genuinely better or merely different.
Why sweet proteins are different
Sweet proteins are proteins that taste extremely sweet, so you only need a tiny amount. In theory the body digests them like any other protein, they don't push blood sugar levels up, and they may even be good for the bacteria in our gut.
Brazzein is the best-known example in the current commercial race. It was originally identified in the fruit of a West African plant, but companies can now produce it using precision fermentation.
In plain English, engineered microbes act as tiny factories. They make the target protein in controlled fermentation tanks, after which the protein is separated and purified.
This is a very different proposition from adding a large quantity of a conventional bulk sweetener. However, sweet proteins aren't a drop-in replacement for sugar.
Sugar does far more than make confectionery sweet. It adds bulk, holds the structure together, helps browning, manages moisture, controls how the product freezes and creates that familiar feel in the mouth. Remove it and the entire product can change.
A successful recipe could therefore combine a sweet protein with fibres, less common sugars such as allulose, starches, fats, flavours and other ingredients that add bulk.
The sweet protein might solve part of the sweetness challenge while the rest of the formulation rebuilds texture and indulgence.
That's the opportunity.
The question isn't whether brazzein can replace sugar one-for-one. It can't. The more valuable question is whether a very small amount of sweet protein can help brands make a much larger sugar reduction without creating a disappointing product.
Oobli and the new sweet protein race
Oobli is the central example because it's taking brazzein beyond ingredient development and into consumer-facing confectionery.
The company produces brazzein through precision fermentation and has launched chocolate products designed to demonstrate what a protein-based sweetening system can do. Oobli reports that US regulators reviewed the safety evidence for three versions of its brazzein and raised no objections, a status called Generally Recognised As Safe, or GRAS, for its intended uses, alongside FEMA GRAS designations for flavour use.
Oobli is showing something important. The ingredient isn't being discussed only as a technical possibility. It's being placed into real products where consumers can judge taste, texture and enjoyment.
Other companies are extending the field.
Amai Proteins has developed sweelin, a monellin-based sweet protein designed using computers to model the protein, combined with precision fermentation and food technology. The company reports that sweelin received a US safety review that raised no objections, known as GRAS, in February 2026 and Singapore approval in May 2026.
Nambawan Spain has commercially launched Thaûma in 2026. It's a thaumatin II sweet protein made through plant molecular farming. That simply means plants are used as biological production systems.
Sweegen's Ultratia brazzein also received US FDA GRAS clearance in 2026.
There are also infrastructure signals.
Bestzyme completed a dedicated sweet-protein production line in China in September 2026. Pentasweet is developing a EUR 65 million facility in Lithuania expected to open in 2027. Earlier work by MycoTechnology on a honey-truffle-derived sweet protein provides useful historical continuity.
These developments haven't reached mass adoption, but they do suggest that the supply chain is becoming more serious.
European market estimates currently place fermented brazzein at around EUR 8,000 to 12,000 per kilogram. That's a market estimate, not a universal price. The economics will depend on production scale, how much the microbes can produce, how efficiently the protein is purified and how much ingredient a finished product needs.
At thousands of times the sweetness intensity of sugar, a high price per kilogram might still be workable in some applications. But the final test will always be cost in use.
The opportunity is a system, not an ingredient
For confectionery leaders, I see five immediate opportunities.
- Rebuild the formulation: Use sweet proteins alongside fibres, fats, starches, less common sugars such as allulose and flavour systems to replace the different jobs sugar does.
- Reduce sweetness more intelligently: Don't automatically recreate the same sweetness level. Use the technology to move towards a more restrained and sophisticated taste profile.
- Design for trust: Explain what the ingredient is, how it's made and what the evidence does and doesn't show.
- Test the full experience: Evaluate attributes like texture, melt, snap, browning, shelf life, flavour release and aftertaste, not just sweetness intensity.
- Create a portfolio strategy: Treat sweet proteins as one tool within a broader reformulation system rather than betting everything on one molecule.
Why AI changes the speed of reformulation
Here's the part that excites me most.
The reason sugar reduction has been so slow isn't only ingredient performance. It's the sheer difficulty of rebuilding a food matrix, the overall structure and mix of a food. Every time sugar is reduced, formulators must rebalance sweetness, bulk, texture, mouthfeel, browning, moisture, shelf life and how the flavour comes through at the same time. That's thousands of interacting variables, and it's traditionally been solved through long, expensive cycles of trial and error.
This is exactly the kind of problem AI is built for.
Machine learning models can predict how ingredients will interact before a single batch is made. They can screen thousands of combinations of sweet proteins, fibres, rare sugars, starches, fats and flavour systems, then point formulators towards the combinations most likely to work. AI can even suggest entirely new blends and process conditions that no human would have thought to test first.
That's a genuine shift in the timeline. Previous sugar replacers advanced through slow, sequential reformulation, one product at a time. AI compresses that cycle dramatically, moving the work from physical trial and error towards computational prediction, with the physical trials reserved for the most promising candidates.
The practical implication for confectionery leaders is significant. A reformulation programme that once took years could shrink to months, and the range of options becomes far wider.
That means more product formats become viable, including premium and indulgent ones that were previously written off as impossible to reduce.
AI won't make the final judgement on taste. Human research and development teams, and consumers, still will. But it can hand formulators a far better starting point than ever before, which changes the economics of sugar reduction entirely.
The commercial opportunity won't come from claiming that sweet proteins are automatically healthy, even though they probably are,. It'll come from making products that use less sugar, deliver a compelling eating experience and are supported by transparent evidence.
Four horizons for confectionery leaders
One year
Manufacturers should audit their sweetener portfolios, claims, exposure levels and consumer trust risk.
That means mapping where sugar-based sweeteners and intense sweeteners such as stevia or sucralose are used, understanding the evidence behind public claims and identifying products that could face reputational pressure.
It also means watching the emerging sweet protein supply base without assuming every announced capacity project will become commercially viable.
One to Three years
Sweet proteins and blended systems may move from demonstrations into targeted confectionery, beverage and dairy applications as production capacity expands.
I expect the first strong use cases to appear where a tiny amount of high-potency sweetness can solve a clear sensory problem. Premium chocolate, functional confectionery, reduced-sugar beverages and selected dairy products could all become testing grounds.
Three to Five years
Reformulation may become a systems-design discipline.
Product developers could combine sweet proteins, fibres, less common sugars, fats, flavours and processing know-how from the start. The goal won't be to find one perfect substitute. It'll be to build a better overall recipe that consumers understand and enjoy.
This broader view aligns with the argument in my earlier article Sweet Proteins: The Secret Weapon for Healthy Aging, while keeping the focus here on category trust and commercial formulation.
Five to 10-plus years
Sweetness could become programmable, personalised and increasingly decoupled from large sugar loads.
Imagine products adjusted for individual taste preferences, metabolism or stage of life. That future could connect with the wider shift beyond single-nutrient thinking explored in Healthspan on a Plate Pt 3: The Death of Nutritionism.
But regulation, cost, evidence and consumer acceptance will determine whether that future arrives.
The bigger confectionery reset
The industry has a choice.
It can keep moving from one sweetener controversy to the next, defending each ingredient after the headlines appear.
Or it can design a more credible future.
That future won't be sugar-free at any cost. It won't be built on exaggerated health claims. It won't assume that consumers only care about a number on the nutrition panel. It'll be built around better products, lower sugar loads, clearer communication and a much more honest explanation of trade-offs.
Sweet proteins could help confectionery rebuild trust. But they won't do it alone. They'll need rigorous evidence, realistic regulation, reliable production and brilliant, probably AI powered, formulation. Most importantly, they'll need brands willing to admit that sweetness is only one part of the experience.
If the next generation of confectionery can deliver pleasure with less sugar, a more satisfying food matrix and a clearer biological story, could the category move from defending indulgence to redesigning it?
Healthy indulgence could become reality!
I think that's a great opportunity.
Read more on the wider transformation of confectionery in The Confectionery Crisis and the Functional Tech-Driven Cure.
To continue the conversation email me at tony@futuristforfood
