🔍 Read the full analysis: What Google And CSIRO’s Work Says About AI And Kelp Restoration on ThorstenMeyerAI.com
TL;DR
Google’s November 12 post describes how CSIRO and collaborators used Google AI tools and Australian giant kelp reference genomes to analyze wild kelp for restoration research. The reference genomes are published in a peer-reviewed journal; the analysis of 600 specimens and its reported findings have not yet been independently documented in a paper or dataset. Whether selectively bred kelp can survive warmer seas remains unknown.
Google said on November 12 that Australia’s national science agency, CSIRO, used AI tools adapted for giant kelp DNA to analyze 600 wild specimens as part of research aimed at restoring Tasmania’s declining kelp forests. The work may help scientists select plants for selective breeding, but it does not show that AI has restored kelp or that bred plants can withstand warmer seas.
The project’s genomic work starts with two Australian giant kelp reference genomes, assembled from specimens from Victoria and Tasmania. The genomes were published in June 2026 in the peer-reviewed journal BMC Genomics, with CSIRO scientists and a Google researcher among the authors. The paper reports that the Australian genomes differ from one another by 0.2% and from a Californian reference by 1.5%.
A local reference matters because researchers use it to compare DNA from other plants. The study authors say the genetic difference between Northern and Southern Hemisphere giant kelp makes relying on the Californian genome alone potentially misleading. Google said its open-source tools, including DeepConsensus and DeepPolisher, helped correct errors in sequencing data used to build the Australian genomes.
Google’s post and accompanying video describe a separate analysis in which a version of DeepVariant was trained to analyze kelp DNA from 600 wild specimens collected at 40 sites along Tasmania’s coast. Google says the analysis found that remaining kelp patches are genetically isolated. That result, unlike the reference-genome research, has not yet been located in a peer-reviewed paper or public dataset, according to the source report. The AI identifies genetic variation; laboratory tests, not the model itself, assess traits such as heat tolerance.
Giant kelp · Tasmania · CSIRO with Google AI
From the reef to the lab and back: where the AI fits
Five steps, as described by CSIRO and Google. The AI works in steps 2 and 3; people do the diving, the lab work and the planting.
1Reef
Collect wild kelp
600 wild specimens from 40 sites along the Tasmanian coast, according to Google.
2Lab + AI
Build a local reference
Long-read DNA plus Google’s DeepConsensus and DeepPolisher: two Australian reference genomes, published in BMC Genomics.
AI step3AI
Read each plant
A version of DeepVariant trained for kelp finds where each specimen’s DNA differs.
AI step4Lab
Choose and cross
Distinct survivors with genetic variation and heat-tolerant traits are crossed. Selective breeding, not genetic modification.
5Reef
Replant and watch
Young kelp is attached to rocky reef in pilot tests. Warm summers show which plants survive.
Sources: Google blog post, Nov 12, 2026, and Google Research video; Scharfenstein et al., BMC Genomics, 2026; CSIRO, 2024, and CSIRO on Google Arts & Culture. The scene is a schematic, not a map or real data.
Graphic: Powered by Thorsten Meyer AI
From DNA Analysis to Reef Trials
Tasmania’s giant kelp decline is the reason researchers are looking for survivors with useful genetic differences. CSIRO says forests have fallen by 95%, with warming, nutrient-poor water and sea urchins among the pressures. The source material cites earlier research estimating a 95–98% loss of kelp cover in eastern Tasmania since 1940; those figures refer to a specific region and historical period, rather than a new measurement announced by Google.
If the genetic analysis helps scientists identify and cross distinct native survivors, it could inform restoration choices. But the work is one step in a longer chain: genomic comparisons can help identify candidates, and laboratory tests can examine their traits, while survival in the ocean remains the key practical test. Until planted kelp is monitored through warm conditions, the project has not established that breeding can reverse the decline.
The stakes extend beyond kelp itself. Giant kelp forests provide habitat and nursery areas for marine life, including fish and invertebrates. Their loss can affect a wider coastal ecosystem. Google’s announcement therefore describes research infrastructure and a possible breeding approach, not a demonstrated ecosystem recovery.
Popular right now
Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Bluetooth 5.3, USB-C Charging, IPX4, Black
Vitamix Touch & Go Advance 32-Ounce Bar Blender, White
Shark WANDVAC Power PET Cordless Handheld Vacuum, White, WV410WH
As an affiliate, we earn on qualifying purchases.
Why Australian Reference Genomes Matter
The restoration project was announced in February 2024 by Google Australia and a group of partners that included CSIRO, the University of Tasmania’s Institute for Marine and Antarctic Studies, The Nature Conservancy, the Great Southern Reef Foundation and the Kelp Forest Alliance. It had two strands: mapping kelp from satellite imagery and studying its genetics. The November announcement concerns the genomics strand.
The mapping strand was planned to use Google Earth Engine and Vertex AI to examine more than 7,000 square kilometres of satellite imagery for remaining kelp. That planned coverage is separate from the reported genetic analysis of specimens gathered at 40 sites. The two Australian reference genomes, meanwhile, address a basic research need: giving scientists a more suitable regional baseline for comparing Australian plants.
Giant kelp, Macrocystis pyrifera, grows in cold, shallow seas. CSIRO says plants in Tasmanian waters can reach about 30 metres. The agency attributes pressure on the forests to warmer, nutrient-poor water and invading sea urchins as the East Australian Current moves farther south. It also says warming in the Tasman Sea is occurring at twice the global average.
„trained a version of our AI model DeepVariant to analyze our wild samples“
— Google, in a video accompanying its November 12 post
The Missing Field-Survival Evidence
The 600-specimen analysis and Google’s claim that remaining Tasmanian patches are genetically isolated are described in the company’s post and video. The source report says it found no paper or dataset documenting those results at publication time. The methods, detailed findings and independent evaluation of that analysis therefore remain unclear from the material available.
It is also not yet clear which specific genetic differences will prove useful for restoration, or how well any selected traits will persist after breeding. The source material says researchers plan to cross distinct native survivors in a laboratory; it describes this as selective breeding, not genetic modification. No result from planting the resulting young kelp on reefs, or from testing their performance through warmer summers, has been reported.
Breeding, Planting and Monitoring
Researchers plan to use the genetic work to choose distinct native kelp for crosses in the laboratory, then plant young kelp on reefs. The next evidence to watch for includes details of the 600-sample analysis, its methods and data, followed by results from breeding and laboratory tests.
The central milestone will be whether planted kelp survives and grows under the warmer conditions affecting Tasmanian waters. The available announcement does not give a timetable for reef planting or monitoring, so when those results will be available is not yet clear.
Key Questions
What did Google and CSIRO use AI to do?
Google says CSIRO and collaborators adapted a version of DeepVariant to analyze DNA from 600 wild giant kelp specimens collected at 40 sites in Tasmania. The AI was used to study genetic variation, not to plant kelp or directly measure heat tolerance.
Has the AI analysis been peer-reviewed?
The two Australian reference genomes were published in peer-reviewed BMC Genomics in June 2026. The source report says it found no paper or dataset yet for Google’s claims about the 600-specimen analysis and genetic isolation.
Does this mean kelp restoration has succeeded?
No. The announcement describes genomic research and planned selective breeding. Whether bred kelp survives after planting on reefs, particularly through warm summers, has not yet been shown.
Is the planned breeding genetic modification?
The source material describes the planned work as selective breeding: crossing distinct native survivors in a laboratory. It says the approach is not genetic modification.
Why are Australian reference genomes useful?
Researchers compare samples against a reference genome. The published study reports a 1.5% difference between the Australian genomes and the Californian reference, compared with 0.2% between the two Australian genomes, supporting the need for a regional reference when studying Australian kelp.
Source: ThorstenMeyerAI.com
