Announcement PDF 22 April 2025 90% INCREASE IN KORSNÄS REE RESOURCE Prospech Limited (ASX: PRS, Prospech or the Company) is pleased to announce a material update to the JORC Code (2012) Inferred Mineral Resource Estimate (MRE) for its Korsnäs Rare Earth Elements (REEs) project in Finland. The updated MRE incorporates data from the Company’s 2024 drilling campaign as well as the completed program of sampling and assaying of historic drill core from drilling conducted in the 1950s, 60s and early 70s. At a selected lower cut-off grade of 0.5% TREO1, the updated Inferred MRE is: INFERRED MINERAL RESOURCE ESTIMATE 13.5 Mt @ 1.02% TREO - lower cut-off grade of 0.5% TREO This updated estimate represents a 90% increase in tonnage compared to the previous MRE, while maintaining a comparable average grade (previously 1.08% TREO). The step-change in scale reflects the availability of a complete set of assay data, a substantially improved geological model that has enhanced continuity and correlation between drill sections and down-dip and the use of a revised bulk density of 2.77 tonnes per cubic metre (previously 2.70), now derived from direct measurement rather than assumption. The result marks a significant advance in the Company’s understanding of the deposit, confirming the scale and quality of the Korsnäs REE system and reinforcing its position as one of the most compelling pre-development rare earth projects in Europe. Table 1. Korsnäs Inferred Mineral Resource Estimate at various TREO cut-offs. 1. TREO = Total Rare Earth Oxides which is the sum of La2O3, CeO2, Pr6O11, Nd2O3, Sm2O3, Eu2O3, Gd2O3, Tb4O7, Dy2O3, Ho2O3, Er2O3, Tm2O3, Yb2O3, Lu2O3 and Y2O3. In addition to the reported Inferred MRE, Prospech has also defined an Exploration Target, located adjacent to and along strike from the current resource area. This target has been outlined based on preliminary geological modelling and is reported as follows: 9 million to 11 million tonnes at 0.9% to 1.1% TREO The potential quantity and grade of this Exploration Target is conceptual in nature. There has been insufficient exploration and metallurgical test work to estimate a Mineral Resource, and it is uncertain whether further exploration will result in the estimation of a Mineral Resource. Prospech Managing Director, Jason Beckton, commented: "The updated Inferred Mineral Resource Estimate at Korsnäs marks a major milestone for Prospech and confirms the emergence of the EU’s newest rare earths resource. With a 90% increase in resource tonnage, now standing at 13.5 million tonnes at 1.02% TREO, and underpinned by a significantly improved geological model and a comprehensive modern assay dataset, this result highlights both the growing scale and strategic potential of the Korsnäs project. It should be noted that the Korsnäs project MRE is not fully defined. As anticipated in our maiden resource announcement, some of the then reported Exploration Target tonnes have now been converted into our increased MRE. Despite the 90% growth in the Korsnäs MRE tonnage, our updated Exploration Target is almost unchanged offering further significant upside potential in more open, under-drilled areas directly along strike from known, mineralised, structures (see Figure 18). The MRE is based on 41 defined mineralised domains, incorporating data from our 2024 diamond drilling program and 4,035 assays from 237 historical drill holes preserved by the Geological Survey of Finland (GTK). In total, 479 historic holes are archived by GTK, of which 275 were relogged and 237 sampled by Prospech geologists. Bulk density has been measured directly, rising from 2.70 to 2.77 tonnes per cubic metre and enhancing confidence in the tonnage estimate and the overall robustness of the model. The next phase of work includes targeted drilling to expand and connect mineralised domains, further metallurgical test work, ongoing economic evaluation and mine permitting activities. With additional drilling, we expect the Exploration Target may be substantially converted to a resource estimate. We also see immediate upside potential from the previously drilled and assayed Tailings Storage Facility (TSF) and Lanthanide Concentrate Stockpile (LnCS) which may provide low-cost feed options for a potential starter operation. With multiple workstreams now underway, we are confident that Korsnäs is advancing towards becoming a significant near-term contributor to the European rare earths supply chain." Background The Korsnäs lead mine, located on Finland's west coast near the Gulf of Bothnia, was operational from 1958 to 1972. During this period, 0.87 million tonnes of ore were processed on-site at a recovered grade of 3.6% lead (Pb). Outokumpu Oy, the original mine operator, was aware that the Korsnäs orebody contained REEs. Although an REE concentrate was produced on-site, it was never sold and remains stockpiled at the mine. Prospech holds 100% of the Korsnäs project which features at least six sub-parallel geological structures up to 20 metres horizontal thickness rich in REEs. Prospech’s exploratory efforts have focused on investigating these structures which are the subject of this resource estimate. 2 Location and Access The Korsnäs REE project is located on Finland's west coast, near the Gulf of Bothnia (Figure 1). It is centred around a historic lead mine, situated 45 km southwest of Vaasa, 4 km northeast of Korsnäs and 171 km from the significant port of Kokkola. The main road between Vaasa and Korsnäs passes through the project area, which has access to grid power and communications. Figure 1. Korsnäs project is located south of the Ostrobothnia capital Vaasa where the potential battery manufacturing facility, Energy Vaasa industrial park, is being established. 3 Tenure Tenement areas are reserved for exploration by Reservation Notifications, with priority going to the first applicant, followed by the granting of an Exploration Permit and then a Mining Permit approved by the Finnish mining authority, the Finnish Safety and Chemicals Agency (TUKES): Tenure at Korsnäs comprises 4 tenements (Figure 2): • • • • ML2021:0019 Hagg1 (182.32 Ha) ML2025:0020 Hägg 23 (185.55 Ha) ML2024:0087 Hägg 32 (167.15 Ha) ML2024:0103 Petalax3 (2,995.37 Ha) 1. Granted by TUKES on 7 May 2024. Granted by TUKES on 10 April 2025. If no appeals lodged to Administrative Court, becomes valid on 19 April 2025. Exploration Permit Applications filed with TUKES for handling and granting of Exploration Permits. 2. 3. The Mining Act (621/2011; Amendments up to 573/2023) (Mining Act) is the relevant legislation which governs exploration and mining activities in Finland. Without the consent of the owner, the Mining Act prohibits the granting of an Exploration Permit within 150 metres of certain building structures (section 7 of the Mining Act). Hence the Exploration Permit exclusion zones (best shown as the ‘bubbles’ in the Petalax Reservation Notification tenement) in the map of the Korsnäs project tenements in Figure 2 below. As previously announced (ASX announcement on 4 October 2023: Korsnäs Municipality Support Expands the Korsnäs Project), the Company has received consent of the Korsnäs Municipality to allow Prospech to conduct exploration activities on areas owned by the Korsnäs Municipality and within the 150 metre Exploration Permit exclusion zones. 4 Figure 2. Korsnäs project tenement map showing ML2021:0019 Hagg (in yellow), ML2025:0020 Hägg 2 and ML2024:0087 Hägg 3 (in orange) and ML2024:0103 Petalax (in green). The Mining Act does not prohibit the granting of a Mining Permit in these same areas. The Mining Act provides for the Government to grant a Mining Permit, known as a ‘redemption permit for a mining area’, over these same exclusion zones if the mining project is based on public need and the mining area meets the requirements laid down in the Mining Act (sections 19 and 20 of the Mining Act). In summary, as demonstrated, exploration activities can be conducted within exclusion zones with the consent of the owner and mining activities can be conducted within exclusion zones in accordance with a redemption permit for a mining area. 5 Surveying and Topographic Control Publicly available LIDAR data from Finnish Government sources was utilised for topographic control. The LIDAR topography closely matched the differential GPS surveys of drill collars, which were conducted by a licenced surveyor from Mitta Oy. Drill hole collars were laid out by Mitta Oy. Diamond Core Drilling Data Historic drilling A total of 479 drill holes have been collated in a modern MS Access relational database. Total metreage included in the estimate is 51,814 metres and average depth of 108 metres. Prospech estimates the value of the previously drilled diamond drill core, almost all of which is preserved, to be in excess of $10 million. Observed carbonatite zones were primarily sampled from BQ and AQ drill cores. Half-core or quarter-core samples were initially processed at GTK's Loppi facility and more recently at Palsatech Oy's facility in Kemi. In total there are currently 4,035 assays records from historical drilling in the resource estimate database which were assayed by ALS in Outokumpu. All geology data critical to mineralisation zone definition has been entered where available. No geological relogging was required and existing lithology codes were adopted. Modern era diamond drilling (2024) In August/September 2024, drill testing (KR305 to KR310 continuing KR series from previous drilling) was conducted by Prospech with 6 drill holes for combined total of 1,032 metres. A total of 341 samples were analysed for various elements, including the lanthanide series REEs. Quarter-core sampling was completed to allow for sufficient modern core (half-core HQ) for metallurgical test work. Figure 3. Location of 2024 drill collars has the dual focus of confirmation of previous grades and providing modern era material for ongoing metallurgical test work. The Tailings Storage Facility (TSF) and Lanthanide Concentrate Stockpile (LnCS) are not included in the Inferred MRE. 6 Sampling and Assaying Samples were quarter cored HQ for the 2024 drilling campaign but also NQ and in some cases BQ for historical sampling. These samples were dispatched by road to the ALS laboratory in Outokumpu. The assay suite, including the ALS methods and detection limits, is detailed in Table 2. ALS method ME-MS81h, which involves lithium borate fusion with ICP-MS analysis, is suitable for ore grade REEs. Additionally, method ME-ICP61 uses a four-acid digestion on a 0.25g sample, followed by ICP-AES analysis was used for other elements. Figure 4. Core photographs are matched with geological logging for both historic and 2024 drill campaign. 7 Table 2. Analytical methods and detection limits. Sample security involved bolted down box dispatch to ALS in Outokumpu. Figure 5. Sample series marked in green with orientation line from which structural information has been collected (314 in total from KR 305 to KR309). Handheld TREO recordings are also shown to ensure sampling of non-obvious mineralised skarn away from obvious, white, carbonatite REE hosting veins. 8 Bulk Density Measurement Procedure and Results Definition Bulk Density is the mass of rock per unit volume, expressed in tonnes per cubic metre (t/m³). It is determined by comparing the weight of the rock sample to the volume of water it displaces, based on Archimedes' principle. Methodology The water displacement method was used to determine Bulk Density. The formula applied is: • Bulk Density (t/m³) = Weight of rock sample (Wrock) / Weight of displaced water (Wwater) In the case of drill core measurements, the following information was recorded for each sample: • • • • • Hole ID Sample depth Sample type (whole, half, or quarter core) Dry weight of the sample Weight of displaced water Results A total of 201 Bulk Density determinations were completed on whole, half and quarter HQ drill core samples from drill holes KR305 to KR310. Of these, 127 measurements are located within the current resource model wireframes. The results (Figure 6) show a normal distribution, with a calculated mean Bulk Density of 2.77 t/m³ adopted as the global Bulk Density estimate for the current MRE. Figure 6 Histogram of Bulk Density measurements. 9 Structural Measurements Where conditions permitted, core orientation was performed on every drill run. Core segments were reassembled and a reference orientation line was drawn. When the orientation line aligned consistently with the markings at both the top and bottom of the run, higher confidence was attributed to the resulting structural measurements. Oriented measurements were collected on metamorphic foliation, joints, faults/shears, veins and lithological contacts. Figure 7 presents plots of selected structural data. Overall, the measurements confirmed that the mineralised zones generally dip shallowly to moderately towards the east, supporting the current geological model and aiding in resource domaining and block modelling. Figure 7. Identifiable white carbonatite about to be measured in the orientation tube. Structural data from orientation tubes as depicted for KR305 to KR307 to stereoscope interpretation. Metallurgical Test Work A comprehensive metallurgical test work program is currently underway at GTK Mintec and the University of Oulu Mining School, as part of the European Commission’s €16.0 million Rare Earth and Magnets Hub for a Resilient Europe (REMHub) initiative. Prospech, through its wholly owned Finnish subsidiary Bambra Oy, is a partner in the REMHub program, an EU funded program focused on strengthening Europe’s access to rare earth elements and permanent magnets, both critical to the green energy transition. The program supports improved supply security through domestic exploration, REE mapping and the development of recycling and reprocessing technologies. Under REMHub, Bambra Oy has secured a €432,250 grant to support exploration and test work at the Korsnäs project. The metallurgical program at GTK Mintec and the University of Oulu aims to develop efficient processing techniques for REE-bearing materials sourced from the project. Participation in REMHub positions Prospech in alignment with the EU’s broader REPowerEU strategy, which seeks to reduce reliance on imported critical raw materials and accelerate the rollout of renewable energy technologies. 10 In parallel with the REMHub metallurgical program, the Company has initiated a complementary metallurgical test work campaign at PT Geoservices’ laboratory in Jakarta. This program involves five composite samples, each weighing approximately 20 to 25 kilograms, prepared from coarse rejects retained from prior assay preparation. The five composites are designed to represent key material sources: • Two samples from 2024 drill core • Two samples from the TSF • One sample from the LnCS As illustrated in Table 3, this work is progressing rapidly and the results will directly inform and complement the metallurgical investigations underway at GTK Mintec and the University of Oulu. The PT Geoservices work provides an early assessment of material variability, mineral deportment and potential beneficiation pathways across both primary and secondary feedstock sources at Korsnäs. Table 3 Progress of PT Geoservices metallurgical program. 11 As a precursor to the REMHub metallurgical program, a detailed mineralogical study was completed by KU Leuven Masters student Niel van de Kerkhof, titled "Investigating the Origin of REE Mineralisation in the Korsnäs Pb-REE Deposit, Finland: Magmatic Carbonatite Dykes or Hydrothermal Veins?" The study incorporated petrographic analysis, mineral chemistry, whole-rock geochemistry and cold cathodoluminescence microscopy. Results confirm that the dominant REE-hosting mineral is fluorapatite (Ca₅(PO₄)₃F), with subordinate bastnäsite, monazite, and minor contributions from allanite, britholite, titanite, and vesuvianite. The REE-bearing carbonatite units are primarily composed of calcite, with accessory phases including pyrrhotite, pyrite, galena, and Ba-orthoclase. This mineralogical framework supports the current metallurgical program’s focus on liberation characteristics, REE deportment and the potential beneficiation pathways for apatite-dominant mineralisation. The predominance of fluorapatite suggests that physical separation (e.g. flotation, magnetic separation) combined with selective leaching may be viable. Understanding mineral associations and grain size relationships is also key to optimising recovery and guiding process flowsheet development. These findings are critical inputs into the ongoing resource evaluation and future upgrade of the Korsnäs resource classification. Figure 8. An example from Van de Kerkhof 2024, depicting A) Unaltered gneiss; B) Altered rock with chalcedony matrix and pink fluorapatite and monazite; C) Altered carbonatite; and D) Altered gneiss with calcite and iron oxides. Figure 9. 10m of core with two carbonatite dikes with skarn on either side of the calcite carbonatites. Importantly these units are easily discerned/geologically modelled and well logged in the past in the case of historic core. 12 Figure 10. Idealised sketch of the Korsnäs carbonatite dykes and accompanying skarn. Calcite interior with sulfides and REE apatite with inclusions of bastnasite and monazite (Van de Kerkhof 2024). QA-QC Duplicate quarter-core samples were inserted into the sample stream to check for variability. Internal ALS laboratory assays of blanks, standards, and duplicates did not reveal any significant issues. A total of 14 duplicate pairs have been analysed from the recent KR305 to KR310 series of drilling. This was not possible from sampling of the historic core due to a requirement to leave behind significant core for future reference. All samples are equal weighting in terms of quarter-core of the exact same metreage and is blind to the ALS laboratory. Table 4. Duplicates completed for KR 305 to KR310 in 2024 campaign indicate acceptable correlation. Dataset will increase to allow statistical correlation in next resource estimate iteration. Sample_ID Sample_DUP Hole_ID From To Ce Dy Dy Dup Er Er Dup Eu Eu Dup Gd Gd Dup Hf Hf Dup Ho Ho Dup La Nd Nd Dup Pr M002624 M002625 KR-305 37.00 38.00 89.0 89.0 4.4 4.5 2.0 1.8 1.5 1.2 6.2 5.8 5.0 4.0 0.7 0.7 44.0 43.0 0.2 0.2 10.0 11.0 38.7 37.7 10.0 9.8 116.0 120.0 7.7 6.8 0.8 0.9 0.2 0.2 21.0 21.0 1.8 1.4 40.0 M002649 M002650 KR-306 61.00 62.00 121.0 103.0 5.2 5.2 2.9 2.6 1.8 1.5 6.0 6.0 5.0 6.0 1.0 1.0 66.0 53.0 0.4 0.4 17.0 17.0 49.9 43.4 13.2 11.5 110.0 108.0 8.5 7.8 1.0 1.0 0.4 0.3 28.0 26.0 2.7 3.0 60.0 50.0 M002674 M002675 KR-306 116.30 116.80 712.0 661.0 12.3 11.7 3.7 4.3 16.0 15.7 31.4 28.9 5.0 4.0 1.7 1.8 335.0 310.0 0.2 0.3 24.0 23.0 379.0 357.0 93.9 87.5 102.0 53.9 54.9 3.3 2.7 0.4 0.4 51.0 48.0 2.3 2.1 260.0 230.0 M002699 M002700 KR-306 178.00 179.00 18150.0 14750.0 27.1 25.3 6.0 5.4 64.6 55.0 113.5 96.1 2.0 2.0 3.3 3.1 10850.0 8820.0 0.4 0.3 95.0 92.0 4920.0 3950.0 1775.0 44.0 384.0 315.0 8.7 7.9 0.6 0.6 68.0 62.0 3.6 3.2 10000.0 8050.0 M002724 M002725 KR-306 246.50 247.50 69.0 74.0 4.2 3.5 1.6 1.4 1.8 1.7 4.7 5.3 4.0 4.0 0.6 0.7 35.0 37.0 0.2 0.2 14.0 15.0 29.5 30.8 7.3 8.4 132.0 135.0 6.9 5.0 0.8 0.7 0.2 0.2 18.0 17.0 1.4 1.4 30.0 30.0 M002749 M002750 KR-307 61.00 62.00 746.0 640.0 4.3 3.4 1.9 1.6 2.2 1.7 7.9 6.8 4.0 4.0 0.7 0.6 508.0 427.0 0.2 0.2 17.0 19.0 188.0 152.0 64.4 52.9 145.0 147.0 16.8 14.0 1.0 0.7 0.2 0.2 20.0 17.0 1.8 1.4 440.0 410.0 M002774 M002775 KR-307 87.00 88.00 124.0 139.0 4.6 3.4 2.6 1.9 1.4 1.7 5.8 5.0 5.0 4.0 0.8 0.6 66.0 82.0 0.3 0.2 11.0 11.0 46.5 46.7 12.4 12.9 163.0 160.0 7.3 7.1 0.9 0.7 0.3 0.3 25.0 17.0 2.2 1.6 60.0 70.0 M002799 M002800 KR-307 113.00 114.00 367.0 187.0 4.9 4.4 2.3 2.2 -0.2 -0.2 7.5 7.2 1.0 2.0 0.9 0.8 267.0 116.0 0.3 0.3 2.0 2.0 98.9 65.7 32.2 18.2 109.0 114.0 11.2 9.4 0.9 0.9 0.3 0.3 29.0 28.0 1.9 1.9 240.0 100.0 M002824 M002825 KR-307 195.00 196.00 84.0 88.0 3.6 4.1 2.1 2.1 1.2 1.1 5.5 5.6 6.0 5.0 0.7 0.7 41.0 44.0 0.3 0.3 11.0 10.0 33.7 34.9 9.3 9.9 136.0 135.0 6.5 5.8 0.7 0.8 0.3 0.3 21.0 21.0 2.0 1.9 30.0 40.0 M002849 M002850 KR-307 219.00 220.00 141.0 166.0 5.2 5.6 2.7 2.8 2.2 2.4 8.2 9.0 5.0 5.0 1.0 1.0 71.0 85.0 0.4 0.3 13.0 14.0 64.2 74.3 16.7 18.9 114.0 110.0 10.8 12.7 1.0 1.1 0.4 0.3 28.0 27.0 2.6 2.8 60.0 70.0 M002874 M002875 KR-308 99.00 100.00 64.0 60.0 4.2 4.2 2.3 2.2 1.0 1.1 5.1 4.7 4.0 5.0 0.9 0.8 31.0 28.0 0.3 0.3 11.0 11.0 27.2 26.1 7.1 6.8 117.0 114.0 5.5 5.1 0.6 0.7 0.4 0.3 23.0 22.0 2.3 2.1 30.0 30.0 M002899 M002900 KR-309 9.00 162.0 192.0 4.0 4.0 1.3 1.2 1.8 1.9 9.2 10.1 4.0 5.0 0.6 0.6 87.0 105.0 0.2 0.1 9.0 6.0 64.7 73.2 18.3 21.7 124.0 128.0 11.9 13.7 1.0 1.0 0.2 0.1 17.0 16.0 0.9 0.8 80.0 90.0 M002924 M002925 KR-309 142.00 143.00 21200.0 23200.0 26.4 26.5 5.4 5.8 82.9 91.3 134.0 138.0 1.0 1.0 3.0 3.1 14250.0 15750.0 0.4 0.5 6.0 1.0 464.0 507.0 10.6 10.6 0.6 0.6 64.0 68.0 3.2 3.8 10000.0 10000.0 M002949 M002950 KR-309 185.50 186.00 469.0 7.1 2.6 2.7 7.2 18.4 16.8 4.0 4.0 1.2 1.1 28.9 2.0 1.7 0.3 0.3 31.0 30.0 1.7 1.6 10.00 577.0 Ce Dup 7.7 7.5 290.0 La Dup Lu Lu Dup Nb Nb Dup 235.0 0.3 13 0.2 25.0 6.0 5920.0 6470.0 2060.0 23.0 245.0 204.0 62.7 Pr Dup 1435.0 2220.0 Rb Rb Dup Sm Sm Dup Tb Tb Dup 39.0 2.0 91.0 50.9 238.0 232.0 33.9 Tm Tm Dup Y Y Dup Yb Yb Dup La 240.0 La Dup 30.0 190.0 14 3D Modelling Multiple mineralised zones were modelled as individual domains, each corresponding to discrete geological structures, primarily carbonatite-hosted lodes. A total of 41 separate domains were interpreted and constructed based on geological continuity, lithological boundaries and assay distribution. These domains were used to control grade interpolation within the block model, with boundaries defined to limit grade smearing and ensure that estimates respected the geometry and distribution of the mineralisation. Upper Bound Estimate: To base glacial till. Lower Bound Estimate: Not greater than 400m below surface. Table 5: 41 domains: Each block in the model has a unique dip and dip direction calculated from the nearest part of the surrounding wireframe. The averages are shown in the table. The 3D modelling process proceeded as follows: 1. Sectional and Flitch Strings: Modelling was carried out using an iterative process involving sets of horizontal and sectional slices and closed strings were digitised for each slice. Geology and structure guided the creation of these shapes which were designed to minimise the inclusion of internal and marginal waste. Confidence in the orientation of the zones was derived not only from sectional interpretation but also from the significant structural database DH_Structural Log - recovered from the logging process and displayed below to illustrate the use of collected data in analysis of the mineralisation controlling vein structures. A total of 314 structural readings were taken from KR305 to KR309. KR310 was vertical and hence not orientated. 15 2. 3D Wireframes: The final set of sectional strings was used to construct the 3D wireframes which were subsequently closed and validated. The wireframes were trimmed at the top to coincide with the base of the glacial till layer, as defined by drilling data, and at depth to exclude previously mined stopes within the historic Korsnäs mine. 3. Compositing: Assays were composited into 1 metre intervals, reflecting the dominant 1 metre sampling interval used in the drilling program. Zero values were inserted into the assay file to ensure that the block model was not inadvertently populated with grade where mineralisation is known to be absent. 4. Block Modelling: A block model was created in Micromine, constrained by 3D wireframes representing the mineralised zones (Table 6). Block percentage factors were applied to accurately account for blocks that lie partially within the wireframes, ensuring a more precise estimation of volume and tonnage. Table 6: Block Model extents and dimensions. 5. Interpolation: Grades were interpolated into the block model using inverse distance squared (ID²) weighting. A three-pass estimation strategy was applied, with progressively expanding search ellipsoids to fill the block model. For the third pass, the maximum search ellipsoid dimensions were set to 180 metres along strike, 180 metres down dip and 6 metres perpendicular to the dip of the mineralised zones. Blocks beyond the defined search ellipsoid semi-axes are not populated with grades, as shown in Figures 11 and 13. Table 7. Block modelling limits and search parameters. For the purposes of reporting all three passes were amalgamated. 16 Figure 11. Oblique view to the NE to illustrate much of the modelled structures have not been populated by blocks as more drilling is required. This figure indicates that significant future resource growth is probable. 17 Figure 12. Drilling data in terms for assays informing the block model are all from recent sampling from 2023 onwards. A total of 485 drill holes are in the database in terms of hard rock drilling (excludes the TSF and LnCS auger drilling). Not all drill holes are included in the block model if too distance from nearby continuous intercepts. It is expected a future estimate with additional drilling will allow more contiguous estimates of the overall tonnage of the parallel zones. 18 Figure 13. A void model was applied to ensure no estimation within modelled mined areas on the main mine trend. Change in dip on sectional interpretations results in no estimation being applied as the change may be due to offsetting faults. 19 Figure 14. Structural measurements resulting from validated orientated core allowed an accurate interpretation of the mineralised zones (veins) to be depicted. In this case, KR307 and KR308 were measured to confirm the orientation of multiple mineralised zones intercepted in each hole. 20 Table 8. All listed REEs were modelled. 6. Block Model Validation The block model was validated through visual inspection on a section-by-section basis, with strong agreement observed between estimated block grades and original drill hole assays. No material inconsistencies were identified in the grade distribution or geometry of mineralisation. In addition, basic statistical comparison was carried out between the composited sample data and the block model values for TREO. As shown in Figure 15, both datasets exhibit an approximately log-normal distribution, which is typical for geochemical data of this nature. • The mean TREO grade is 3,438 ppm in blocks and 3,691 ppm in samples. • The sample dataset exhibits higher variance (39,349,085) compared to the blocks (17,354,301), reflecting the smoothing effect of inverse distance squared (ID²) interpolation. • The block model shows a narrower distribution, as expected from grade estimation and no signs of over-smoothing or bias are apparent. This analysis confirms that the block model appropriately represents the underlying sample population and is suitable for reporting of an Inferred MRE. 21 Figure 15: Histograms comparing TREO (ppm) values for block model vs. composited samples, both plotted on log scale. KORSNÄS PROJECT INFERRED MINERAL RESOURCE ESTIMATE 13.5 Mt @ 1.02% TREO - lower cut-off grade of 0.5% TREO Table 9. Korsnäs Inferred Mineral Resource Estimate at various TREO cut-offs. 22 Grade Tonnage Relationship Figure 16 depicts the TREO grade-tonnage curve for various TREO cut-off grades: Figure 16. Modelled grade-tonnage curve. Korsnäs Project Inferred Mineral Resource Estimate - Comprehensive list of REE elements Table 10: Estimate of all REE elements. 23 Comparison with Previous Estimate The updated Inferred Mineral Resource Estimate (April 2025) stands at 13.5 million tonnes @ 1.02% TREO using a 0.5% TREO cut-off grade. This represents a substantial increase from the November 2024 estimate of 7.1 million tonnes @ 1.08% TREO at the same cut-off grade. The primary factors contributing to this 90% increase in tonnage are: 1. Completion of a comprehensive assay dataset, incorporating results from Prospech’s 2024 drilling program as well as systematic resampling of preserved historical core. 2. A significantly refined geological model, enabling improved correlation of mineralised structures between sections and enhanced down-dip continuity. This has allowed for more confident domaining and expanded the modelled extents of mineralisation. Figure 17: Comparison with the previous Inferred MRE. Reasonable Potential of Eventual Economic Extraction The prospects for eventual economic extraction of rare earth oxides from the 41 modelled grade domains representing mineralised carbonatite veins at Korsnäs have been assessed by the Competent Person, with consideration given to several material factors including, amongst other matters, the prevailing and forecast rare earth market dynamics, particularly within the EU, the potential influence of geopolitical conditions on REE pricing and, given flatter dips confirmed by detailed structural studies completed by Prospech geologists based on the results from the recent drilling campaign completed by the Company, the potential for a partial open pit mining operation at Korsnäs. While mining methods, processing costs, metal prices and metallurgical recoveries remain uncertain at this stage, the Competent Person has drawn on reasonable assumptions supported by public information from