Mr. Michel Fontaine reports

MACHINE LEARNING BREAKTHROUGH: WINDFALL GEOTEK OUTLINES 1.91 KM EXTENSION SCANDIUM TARGET AT CANADA'S FLAGSHIP CRATER LAKE PROPERTY

Windfall Geotek Inc. has completed the modelling and extraction of the geophysical digital signature for the Scandium Crater Lake deposit for Scandium Canada Ltd. (see press release dated April 3, 2025).

Utilizing its proprietary artificial intelligence platform, Windfall has completed an extensive computational analysis of entire publicly data sets to fingerprint the signature to its Scandium TG zone Crater Lake deposit in the Canadian province of Quebec.

Technical summary and methodology

The project focused on the Scandium TG zone. Key technical highlights include:

• Data integration: The AI system synthesized multidimensional public data sets, including high-resolution magnetic, radiometric and topography data, alongside available drill hole assays and surface rock samples geochemistry.

Signature extraction: By isolating the specific geological descriptors of the Crater Lake scandium mineralization, the company has established a high-fidelity digital discovery pattern.

Target generation: This proprietary signature was deployed across the entire regional land package to identify high-probability anomalies. The system did not find any scandium target outside of the Scandium Canada claims.

The power of digital exploration

Using its proprietary AI system in combination with a strong historical exploration database and SIGEOM data covering 8,803 square kilometres. Windfall Geotek identified three high-priority Scandium extension target zones, including one of 1.91 kilometres and two extensions of the Scandium TG deposit (0.44 and 0.53 kilometre), confirming exactly where the company will drill in the next weeks.

Windfall's AI system analyzed more than 2,475 historical scandium-bearing assays (mostly in the Crater Lake deposit) using advanced machine learning algorithms designed to recognize geological patterns and contexts associated with mineralization. In total, 3,529,702 cells measuring 50 metres by 50 metres were evaluated by the Windfall AI system. This data-driven approach significantly decreases exploration time and costs when compared with traditional exploration methodologies.

Management commentary

Michel Fontaine, president and chief executive officer of Windfall Geotek, commented: "Windfall Geotek's AI-driven targeting platform achieved a 94-per-cent accuracy rate in detecting all historically known scandium occurrences at Crater Lake, confirming the strength of its predictive model. Importantly, no equivalent signatures were identified elsewhere on the area of interest cover by the geophysics (8,803 square kilometres) highlighting the rarity and strategic significance of the deposit. By analyzing over 8,803 square kilometres of data and narrowing the search area by 99 per cent, our technology has isolated a high-confidence digital signature directly aligned with the known mineralization. The absence of similar signatures elsewhere reinforces both the uniqueness of the deposit and the potential for a significant scandium discovery."

Technical details

The technical information presented in this news release has been reviewed and approved by Grigor Heba, chief geologist of Windfall Geotek, a qualified person as defined by National Instrument 43-101.

About Windfall Geotek Inc.

Windfall Geotek is a global leader in AI-driven mineral exploration, turning complex geoscience data into high-potential discoveries with unmatched speed and accuracy. With a multidisciplinary team of geophysicists, geologists and AI specialists, Windfall is validating selected AI-generated targets to demonstrate the reliability of its system. Beyond mining, the company is expanding its expertise into life-saving applications, including landmine and UXO detection.

We seek Safe Harbor.